ChemWiki - User contributions [en]

GFP : Green Fluorescent Protein

2009-09-21T20:53:41Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum. So use the script of David in order to make some frequency calculation !!!

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

successful

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Script of David which enables you to run the calculation in the window and just print the geometry and the frequency in the output. Run as an other jobscript : qsub jobscript_live.sh (change the extension .log in .sh, I changed in order to be sure that nothing will be change in the file).
[[Media:jobscript_live.log]]

Summarize : run again the gfp_water_charges_complete_opt_2.com with the indications that I gave.

Then so as to run some frequency calculation, use the script of David with Freq keyword, and if you want to run some frequency calculation on the model, use the IOp(1/33=1) keyword in order to print the model geometry and then run the calculation with the David jobscript on this geometry.

The different output are big, so I can not put them on the wiki, if you need some specific email me.

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

GFP : Green Fluorescent Protein

2009-09-21T20:52:30Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum. So use the script of David in order to make some frequency calculation !!!

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

successful

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Script of David which enables you to run the calculation in the window and just print the geometry and the frequency in the output. Run as an other jobscript : qsub jobscript_live.sh (change the extension .log in .sh, I changed in order to be sure that nothing will be change in the file).
[[Media:jobscript_live.log]]

Summarize : run again the gfp_water_charges_complete_opt_2.com with the indications that I gave.

Then so as to run some frequency calculation, use the script of David with Freq keyword, and if you want to run some frequency calculation on the model, use the IOp(1/33=1) keyword in order to print the model geometry and then run the calculation with the David jobscript on this geometry.

The different output are big, so I can not put them on the wiki, if you need some specific email me.

<h4>[[Mike : help]]</h4>

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

Mike : help

2009-09-21T20:52:18Z

Alasoro: New page: I put the output here : Media:gfp_water_charges_complete_opt_4.log

I put the output here : [[Media:gfp_water_charges_complete_opt_4.log]]

GFP : Green Fluorescent Protein

2009-09-21T20:52:08Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum. So use the script of David in order to make some frequency calculation !!!

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

successful

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Script of David which enables you to run the calculation in the window and just print the geometry and the frequency in the output. Run as an other jobscript : qsub jobscript_live.sh (change the extension .log in .sh, I changed in order to be sure that nothing will be change in the file).
[[Media:jobscript_live.log]]

Summarize : run again the gfp_water_charges_complete_opt_2.com with the indications that I gave.

Then so as to run some frequency calculation, use the script of David with Freq keyword, and if you want to run some frequency calculation on the model, use the IOp(1/33=1) keyword in order to print the model geometry and then run the calculation with the David jobscript on this geometry.

The different output are big, so I can not put them on the wiki, if you need some specific email me.

<h4>[[Mike : help]]</h4>
I put the output here : [[Media:gfp_water_charges_complete_opt_4.log]]

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

GFP : Green Fluorescent Protein

2009-09-21T20:47:39Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum. So use the script of David in order to make some frequency calculation !!!

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

successful

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Script of David which enables you to run the calculation in the window and just print the geometry and the frequency in the output. Run as an other jobscript : qsub jobscript_live.sh (change the extension .log in .sh, I changed in order to be sure that nothing will be change in the file).
[[Media:jobscript_live.log]]

Summarize : run again the gfp_water_charges_complete_opt_2.com with the indications that I gave.

Then so as to run some frequency calculation, use the script of David with Freq keyword, and if you want to run some frequency calculation on the model, use the IOp(1/33=1) keyword in order to print the model geometry and then run the calculation with the David jobscript on this geometry.

The different output are big, so I can not put them on the wiki, if you need some specific email me.

<h4>Mike : help</h4>
I put the output here : [[Media:gfp_water_charges_complete_opt_4.log]]

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

GFP : Green Fluorescent Protein

2009-08-23T21:58:03Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T21:56:29Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum. So use the script of David in order to make some frequency calculation !!!

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

successful

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Script of David which enables you to run the calculation in the window and just print the geometry and the frequency in the output. Run as an other jobscript : qsub jobscript_live.sh (change the extension .log in .sh, I changed in order to be sure that nothing will be change in the file).
[[Media:jobscript_live.log]]

Summarize : run again the gfp_water_charges_complete_opt_2.com with the indications that I gave.

Then so as to run some frequency calculation, use the script of David with Freq keyword, and if you want to run some frequency calculation on the model, use the IOp(1/33=1) keyword in order to print the model geometry and then run the calculation with the David jobscript on this geometry.

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

GFP : Green Fluorescent Protein

2009-08-23T21:50:09Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum.

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

successful

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Script of David which enables you to run the calculation in the window and just print the geometry and the frequency in the output. Run as an other jobscript : qsub jobscript_live.sh (change the extension .log in .sh, I changed in order to be sure that nothing will be change in the file).
[[Media:jobscript_live.log]]

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

File:Jobscript live.log

2009-08-23T21:45:24Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T21:45:09Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red.com : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum.

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Script of David which enables you to run the calculation in the window and just print the geometry and the frequency in the output. Run as an other jobscript : qsub jobscript_live.sh (change the extension .log in .sh, I changed in order to be sure that nothing will be change in the file).
[[Media:jobscript_live.log]]

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

File:Gfp nowater charges complete opt model 1 freq 2.gjf

2009-08-23T21:39:38Z

Alasoro:

File:Gfp nowater charges complete opt model 1 freq 1.gjf

2009-08-23T21:38:51Z

Alasoro:

File:Gfp nowater charges complete opt model 1.gjf

2009-08-23T21:38:22Z

Alasoro:

File:Gfp nowater charges complete opt 1 freq 1.gjf

2009-08-23T21:37:17Z

Alasoro:

File:Gfp nowater charges complete opt 1.gjf

2009-08-23T21:36:49Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T21:34:23Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]]
* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]]
* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... so we need to run this calculation (use the input below, but with a memory of 14000mb and a minimum of 500h of time job)

Input, Output:[[Media:input_opt_2.gjf]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

[[Media:gfp_nowater_charges_complete_opt_1.gjf]]

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry

[[Media:gfp_nowater_charges_complete_opt_1_freq_1.gjf]]

No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red.com : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum.

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

[[Media:gfp_nowater_charges_complete_opt_model_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_1.gjf]]

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_nowater_charges_complete_opt_model_1_freq_2.gjf]]

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

GFP : Green Fluorescent Protein

2009-08-23T21:23:34Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]][[Media:output_spe_1.log]]

* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]] [[Media:output_opt_1.log]]

* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]] [[Media:output_freq_1.log]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]] [[Media:output_spe_2.log]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... we'll see, to complete

Input, Output:[[Media:input_opt_2.gjf]] [[Media:output_opt_2.log]]

[[Media:gfp_water_charges_complete.zip]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry
No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red.com : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red.com : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum.

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

File:Gfp last test.xls

2009-08-23T21:18:57Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T21:18:25Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]][[Media:output_spe_1.log]]

* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]] [[Media:output_opt_1.log]]

* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]] [[Media:output_freq_1.log]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]] [[Media:output_spe_2.log]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... we'll see, to complete

Input, Output:[[Media:input_opt_2.gjf]] [[Media:output_opt_2.log]]

[[Media:gfp_water_charges_complete.zip]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry
No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red.com : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red.com : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum.

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

[[Media:gfp_last_test.xls]]

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

GFP : Green Fluorescent Protein

2009-08-23T21:05:11Z

Alasoro:

<h2>[[GFP : introduction]]</h2>

<h2>PDB : Protein Data Bank</h2>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

The one that I will use in the example below will be known with the PDB code : 1W7S .

<h2>Files and different steps</h2>

<h4>Just after PDB file</h4>
Be careful, concerning the files below, the charges of the chromophore are missing !!!!

<p>
After open with GaussView (4 !!!!!), and some changes, calculation can be run on the files :<br/>
- with water around (Hydrogens not in the right places in the space because Gaussview added them randomly) : [[Media:gfp_water_1w7s_ready_to_use.gjf]] <br/>
- no water around, just the protein : chromophore in the High level (hydrogens of the protein added randomly by GaussView too...) : [[Media:gfp_nowater_1w7s_ready_to_use.gjf]] <br/>
</p>

<h4>After charges problem </h4>

Once charges problem solved, I got 2 files, one with the water and one without. In order to compute some test calculation take without water, but in the case of the GFP, if you want to interpret the results take with.

* without water : [[Media:gfp_nowater_charges.gjf]]

* with water : [[Media:gfp_water_charges.gjf]]

<h4>With additional amino acid in the high region (and 1 water molecule)</h4>

Contain all the different amino acid, and the water that Jasper would like to include in the high level calculation : [[Media:gfp_water_charges_amino_acid.gjf]]

<h4>Optimisation </h4>
<p>
First I run optimisations with the file without water : optimisation = no problem, frequency calculation = problem...
<br/>
Once the optimised geometry found I checked if there was some imaginary frequency, so I ran with the keyword Freq=Modelmodes which enables me to get a file with a reasonable size and the information that I want to get. I was lucky because I had no imaginary frequency, but I just got 40 modes with the keyword Freq=ModelModes, instead of 3*36-6 = 102, so there is a problem. <br/>
I imagined two possibilities : because the calculation was very big it crashed when I wanted to print the final output even if I restrained what it had to print. Or there is a limitation in he code and so just the 40 last modes can be printed.<br/>
So I tried to run an other calculation with the optimised geometry by asking the geometry of the model with IOp(1/33=1), once the file obtained, I had a look at the output and the geometry of the model was behind the part below :

<pre>
At end of L120:
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
1 7 6 23.050591 3.469428 -1.008981

</pre>

So that was strange because when I did this previously (using IOp(1/33=1)) I got the atoms outside the model with a -1, and here I have for the atoms outside the model 6 and for the model some big number(as a code), as below

<pre>
---------------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------------------------
.
.
.
.
473 6 20001005 -0.375481 1.652856 0.482589
474 7 20001016 -0.273771 1.012963 1.610015
475 6 20001005 0.594142 -0.055349 1.376775
476 6 20001005 1.027366 -0.007835 -0.040053
477 8 20001024 1.707451 -0.776975 -0.709324
478 7 20001018 0.428530 1.145189 -0.536877
479 6 20001001 0.563225 1.567653 -1.931383
</pre>

So just take the geometry and hope for a frequency calculation on the model did not work, so I took the geometry and delete all the atoms which did not correspond to the model system, and replace the big code number by 0.

</p>

<h4>Calculations tried, run, failed ...</h4>
<p>
I tried to run different kind of calculations, some need to be run again because of a time problem (and some were killed by Simon Burbidge). So I am going to make a list of the calculation with some explanations and if necessary input and output.
</p>

* gfp_water_charges_complete_1.com : Chromophore well defined (hydrogens, bonds... OK), with the water environment
Single point energy successful

Input, Output :[[Media:input_spe_1.gjf]][[Media:output_spe_1.log]]

* gfp_water_charges_complete_opt_1.com : Optimisation of the previous system.
Optimisation successful

Input, Output :[[Media:input_opt_1.gjf]] [[Media:output_opt_1.log]]

* gfp_water_charges_complete_opt_1_freq_1.com : Frequency calculation of the previous optimisation

Input, Output : [[Media:input_freq_1.gjf]] [[Media:output_freq_1.log]]

But one imaginay frequency, so, not a minimum but a transition state

* gfp_water_charges_complete_2.com : Chromophore well defined, with the water environment and the amino acid that interest Jasper (with ONE water molecule) in high level of accuracy.
Single point energy calculation successful

Input, Output :[[Media:input_spe_2.gjf]] [[Media:output_spe_2.log]]

* gfp_water_charges_complete_opt_2.com : Optimisation of the previous system
Problem of job killed.... because I put to many hours... we'll see, to complete

Input, Output:[[Media:input_opt_2.gjf]] [[Media:output_opt_2.log]]

[[Media:gfp_water_charges_complete.zip]]

* gfp_nowater_charges_complete_energy_1.com : chromophore well defined and just chromophore in the high level.
single point energy successful.

* gfp_nowater_charges_complete_energy_2.com : test of oniom(b3lyp/6-31g(d):amber=hardfirst)=scalecharge=555500 (no change concerning this and embedcharge)

* gfp_nowater_charges_complete_opt_1.com : optimisation of the previous system
opt successful

* gfp_nowater_charges_complete_opt_1_freq_1.com : frequency calculation of the previous optimised geometry
No imaginary frequency, but this one is with ModelModes and I just have 40 modes which is really less than what is expected, thus I run again the frequency calculation with a script of david which enable to run the calculation in the terminal shell and create an output file with just the geometry and the frequency calculation inside (that enables us to create a smaller output).

* gfp_nowater_charges_complete_opt_1_freq_1_red.com : ModelModes but with the script of david, because the question was concerning the previsous calculation : I had just 40 modes because that the maximum that can keep in memory gaussian in this such kind of size file, or it is because of a strange option inside the ModelModes, which limited the number of modes in the output

* gfp_nowater_charges_complete_opt_1_freq_2_red.com : just freq calculation, and no imaginary frequency so that is ok ! that is a minimum.

* gfp_nowater_charges_complete_opt_model_1.com : optimisation with the option of printing the geometry of the model system

* gfp_nowater_charges_complete_opt_model_1_freq_1.com : frequency calculation of the model geometry, but failed because of the "page layout", so gaussian did not recognize this way of present the geometry.

* gfp_nowater_charges_complete_opt_model_1_freq_2.com : I changed the different columns of the optimised geometry and run again a frequency calculation and I got many imaginary frequencies (9 imaginary freq)

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

GFP : Green Fluorescent Protein

2009-08-23T20:30:03Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T20:06:05Z

Alasoro:

File:Input opt 2.gjf

2009-08-23T19:58:32Z

Alasoro:

File:Input spe 2.gjf

2009-08-23T19:58:08Z

Alasoro:

File:Input freq 1.gjf

2009-08-23T19:57:31Z

Alasoro:

File:Input opt 1.gjf

2009-08-23T19:56:39Z

Alasoro:

File:Input spe 1.gjf

2009-08-23T19:56:13Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T19:55:27Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T19:51:06Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T16:21:21Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-23T16:11:27Z

Alasoro:

File:Select oniom colum.png

2009-08-13T15:57:19Z

Alasoro:

First steps with ONIOM

2009-08-13T15:56:55Z

Alasoro:

== Create the first ONIOM molecule ==

<p>
Before going further, try to run our first ONIOM calculation !
</p>

<p>
In this aim, take Benzene and distort it, by putting a carbon out of the plan. Then select the two different areas of ONIOM calculation (High and Low level).
First of all draw the molecule, and change a dihedral angle to put a carbon out of the plan.
</p>

[[Image:essai1_creation.jpg]]
<p>
Then select the two different parts of the molecule on which you want to apply different level of theory.
</p>

[[Image:essai1_choixbase.jpg]]
<p>
And finally, choose the basis set for this two different layer by selecting ''Multilayer ONIOM model'' in the '''Calculate ---> Gaussian''' part.

</p>

== Input ==

<p>
You must have an input like this one :
</p>

[[Image:essai1_input_correct.jpg]]

<p>
[[Media:oniom_essai1.gjf]]
</p>

== Output ==

<p>
In the input file we could find "the energy" of the molecule. In fact you could find the 3 different parts of the energy of the molecule, which correspond to this equation :
</p>

<p>
<math> E^{ONIOM} = E_{model}^{high} + E_{real}^{low} - E_{model}^{low} </math>
</p>

<pre>
ONIOM: calculating energy.
ONIOM: gridpoint 1 method: low system: model energy: -115.814040178590
ONIOM: gridpoint 2 method: high system: model energy: -116.609098705187
ONIOM: gridpoint 3 method: low system: real energy: -229.419112698610
ONIOM: extrapolated energy = -230.214171225207
</pre>

<p>
So the <math> E^{ONIOM} </math> corresponds to ''ONIOM: extrapolated energy''
</p>

[[Media:oniom_essai1.log]]

== Mistakes ==

<p>
You just run our first ONIOM calculation, but this calculation is not a good example of what ONIOM can do and for what ONIOM is useful, because the way that we delimited the system between the high and low level, broke the aromaticity of the benzene.
</p>

<p>
The aim of ONIOM is more to consider as a low-level of theory the environment of a definite molecule, which is consider as a high level of theory.
</p>

== An other way to select layer ==

In order to select different layers you can use too the atoms list. I recommend this way specially when you are in the case of a big molecule (like a protein), and that you want to select your chromophore or such specific amino acid that you would like to include in your high level of theory for example.

So Open the atom list, and then click on the L on the top of the list in order to make appear the oniom layers of the atoms. Then click on each layer of the atoms and you will be able to change the layer of the atom.
[[Image:select_oniom_colum.png]]

Get back to [[Introduction]]

Next part [[ONIOM for excited states]]

GFP : Green Fluorescent Protein

2009-08-13T14:14:28Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-13T13:58:59Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-13T10:55:26Z

Alasoro:

Symmetry Lab

2009-08-11T08:41:45Z

Alasoro:

== Hand-in Dates ==

TBC - ideally via e-mail address

== Timetable ==

(to update for 2009 - 2010)]

In ground floor computer room, and also 232A study area.
2 group of about 65 students each: 4 afternoons, 2-5pm.
PhD demonstrators (shown like this).

'''Group B''' <br>
Friday 24th October (Stephen Bourne) <br>
Thursday 30th (Yao Wong + Rob Felstead) <br>
Friday 31st (Stephen Hodge) <br>
Friday 7th November (Stephen Bourne + Michael Inkpen) <br>

'''Group A''' <br>
''Introduction: LTC, Tuesday 18th November, 2pm'' <br>
Friday 21st November (Stephen Hodge) <br>
Thursday 27th (Michael Inkpen + Tim Wilson) <br>
Friday 28th (Rob Felstead) <br>
Friday 5th December (Yao Wong) <br>

----

----
== Exercise ==

<h3>Molecular Symmetry : introduction to the lab, objectives</h3>

<h4>Introduction</h4>
<p>
The exercises run in the computer labs using '''Gaussview''' for Windows.<br/>
(Programs menu).
</p>

<p>
The files you need are available in two places:<br/>
* [https://www.ch.ic.ac.uk/wiki/index.php/Symmetry_Lab_Downloads Symmetry Lab Downloads] <br/>
zip archives for ex1 and ex2: extract the files to a folder in your own directory <br/>
* in the Mac-PC/symmlab folder.<br/>
Use '''‘add a network place->choose another network location’''' to connect to the \\chnts1.ch.ic.ac.uk\mac-pc server first.
</p>

<p>
Copy the folders ex1 and ex2 to your network home directory (or somewhere else you can write to) first. The files contained in these directories should be opened from within '''Gaussview''': you will not be able to open them by double-clicking on a PC.
</p>

<p>
As part of the exercise, you will need to copy images from the Gaussview display. Two suggested ways of doing this: <br/>
* Alt-Print Screen copies the current window to the clipboard. This can be pasted into e.g. PowerPoint or Word. <br/>
* '''Gaussview''' can save graphics files in TIFF, JPEG or BMP format, which can be imported into another application. JPEG is recommended, as the files are compressed and smaller than TIFF or BMP. <br/>

Find a method that works for you before starting your write-up.
</p>

<p>
For exercise 1, you’ll be opening '''Gaussian 09''' input files (myfile.com) and viewing molecular structures. Some instructions and suggestions for using '''Gaussview''' are included in the text. There is also a built-in help menu.
</p>

<p>
For exercise 2, you’ll be opening '''Gaussian 09''' formatted checkpoint files (myfile.fchk) and viewing molecular orbitals and vibrations.

</p>

<h4>Background reading</h4>
<p>
''Atkins and de Paula'', Physical Chemistry 7ed, Chapter 15 (Molecular Symmetry) and references cited therein. Some websites may also be useful:<br/>

[http://www.ch.ic.ac.uk/local/symmetry A tutorial about the subject]<br/>

[http://www.chemistry.nmsu.edu/studntres/chem639/cgi-bin/group1.cgi Group theory tables]<br/>
</p>

<h4>Objectives</h4>

<p>
This ‘experiment’ consists of two exercises. Allow roughly equal time for each.
</p>
<p>
'''Exercise 1''': Determine the symmetry operations that can be applied to several molecules, and hence the molecular point group.
</p>

<p>
'''Exercise 2''': Determine the symmetry properties of some molecular vibrations and molecular orbitals. (This may not have been covered in lectures yet).
</p>

----
<p>
'''At the end''' <br/>
You should be able to <br/>
* Determine the point group of any molecule <br/>
* Determine which irreducible representation of a point group labels the symmetry of a particular molecular vibration or molecular orbital
</p>
----

<p>
You will also have a better understanding of symmetry and some of its applications in chemistry.
</p>

<h4>Concerning the report</h4>
<h5>Mark page, Data sheet</h5>

<p>
Look at the marking scheme carefully, and do not forget to join it to your report : [[Media:Mark_page.doc]]
</p>

<p>
Data sheets available here : [[Media:DATA SHEET_ex1_ex2.doc]] <br/>
You might find it convenient to enter data and make sketches and notes in the data sheets at points indicated in the text by
'''==> Complete data sheet'''
</p>

<h5>Report</h5>
<p>
Suggested guide/template for write up 10-15 pages
You should start by reading the discussions in e.g. Atkins and the handout itself.
Marks are given for correct results, but also for demonstrating that you understand how the results are obtained, and what they mean. A few carefully constructed diagrams with clear explanations are more likely to demonstrate this. Marks will be lost for unthinking copy/paste.
</p>

<h6>Exercise 1:</h6>
<p>
'''Theory''' <br/>
Briefly discuss the concepts of symmetry elements and symmetry operations, groups of symmetry operations (point groups) in your own words.
</p>

<p>
'''Results and Discussion''' <br/>
For the six molecules being tested, give the point group, flowchart assignment, and additional symmetry elements (from the data sheet). All symmetry elements should be illustrated for each molecule.
Comment on the symmetry elements that are present in both cyclohexane (demo) and chlorocubane, and those that are only present in one or other of the two molecules.
</p>
<h6>Exercise 2:</h6>

<p>
'''Theory''' <br/>
Briefly discuss the concept of an irreducible representation of a point group.
</p>

<p>
'''Results and Discussion''' <br/>
Tabulate and summarise the results you collected in completing the data sheet for each of the four examples.
What is a degenerate irreducible representation, and what symmetry element(s) must be present for these to occur? Discuss this, referring to the results you obtained for <math>BFBr_2</math> and <math>BCl_3</math>.
For <math>CrCO_6</math>, discuss the relationship between the correlation diagram, the orbitals you obtained (data sheet / diagrams) and their bonding characteristics. E.g. the <math>T_{2g}</math> orbitals are nonbonding and located mainly on the central metal atom.
</p>

<h6>Conclusions</h6>

<h3>[[Exercise 1 : Point groups - symmetry elements and operations]]</h3>

<h3>[[Exercise 2 : Symmetry properties of molecular vibrations and molecular orbitals]]</h3>

File:DATA SHEET ex1 ex2.doc

2009-08-11T08:41:33Z

Alasoro:

File:Mark page.doc

2009-08-11T08:41:20Z

Alasoro:

Symmetry Lab

2009-08-11T08:41:05Z

Alasoro:

Symmetry Lab

2009-08-11T08:39:03Z

Alasoro:

== Hand-in Dates ==

TBC - ideally via e-mail address

== Timetable ==

(to update for 2009 - 2010)]

In ground floor computer room, and also 232A study area.
2 group of about 65 students each: 4 afternoons, 2-5pm.
PhD demonstrators (shown like this).

'''Group B''' <br>
Friday 24th October (Stephen Bourne) <br>
Thursday 30th (Yao Wong + Rob Felstead) <br>
Friday 31st (Stephen Hodge) <br>
Friday 7th November (Stephen Bourne + Michael Inkpen) <br>

'''Group A''' <br>
''Introduction: LTC, Tuesday 18th November, 2pm'' <br>
Friday 21st November (Stephen Hodge) <br>
Thursday 27th (Michael Inkpen + Tim Wilson) <br>
Friday 28th (Rob Felstead) <br>
Friday 5th December (Yao Wong) <br>

----

----
== Exercise ==

<h3>Molecular Symmetry : introduction to the lab, objectives</h3>

<h4>Introduction</h4>
<p>
The exercises run in the computer labs using '''Gaussview''' for Windows.<br/>
(Programs menu).
</p>

<p>
The files you need are available in two places:<br/>
* [https://www.ch.ic.ac.uk/wiki/index.php/Symmetry_Lab_Downloads Symmetry Lab Downloads] <br/>
zip archives for ex1 and ex2: extract the files to a folder in your own directory <br/>
* in the Mac-PC/symmlab folder.<br/>
Use '''‘add a network place->choose another network location’''' to connect to the \\chnts1.ch.ic.ac.uk\mac-pc server first.
</p>

<p>
Copy the folders ex1 and ex2 to your network home directory (or somewhere else you can write to) first. The files contained in these directories should be opened from within '''Gaussview''': you will not be able to open them by double-clicking on a PC.
</p>

<p>
As part of the exercise, you will need to copy images from the Gaussview display. Two suggested ways of doing this: <br/>
* Alt-Print Screen copies the current window to the clipboard. This can be pasted into e.g. PowerPoint or Word. <br/>
* '''Gaussview''' can save graphics files in TIFF, JPEG or BMP format, which can be imported into another application. JPEG is recommended, as the files are compressed and smaller than TIFF or BMP. <br/>

Find a method that works for you before starting your write-up.
</p>

<p>
For exercise 1, you’ll be opening '''Gaussian 09''' input files (myfile.com) and viewing molecular structures. Some instructions and suggestions for using '''Gaussview''' are included in the text. There is also a built-in help menu.
</p>

<p>
For exercise 2, you’ll be opening '''Gaussian 09''' formatted checkpoint files (myfile.fchk) and viewing molecular orbitals and vibrations.

</p>

<h4>Background reading</h4>
<p>
''Atkins and de Paula'', Physical Chemistry 7ed, Chapter 15 (Molecular Symmetry) and references cited therein. Some websites may also be useful:<br/>

[http://www.ch.ic.ac.uk/local/symmetry A tutorial about the subject]<br/>

[http://www.chemistry.nmsu.edu/studntres/chem639/cgi-bin/group1.cgi Group theory tables]<br/>
</p>

<h4>Objectives</h4>

<p>
This ‘experiment’ consists of two exercises. Allow roughly equal time for each.
</p>
<p>
'''Exercise 1''': Determine the symmetry operations that can be applied to several molecules, and hence the molecular point group.
</p>

<p>
'''Exercise 2''': Determine the symmetry properties of some molecular vibrations and molecular orbitals. (This may not have been covered in lectures yet).
</p>

----
<p>
'''At the end''' <br/>
You should be able to <br/>
* Determine the point group of any molecule <br/>
* Determine which irreducible representation of a point group labels the symmetry of a particular molecular vibration or molecular orbital
</p>
----

<p>
You will also have a better understanding of symmetry and some of its applications in chemistry.
</p>

<h4>Concerning the report</h4>
<h5>Mark page, Data sheet</h5>

<p>
Look at the marking scheme carefully.
</p>

<p>
Data sheets available here : <br/>
You might find it convenient to enter data and make sketches and notes in the data sheets at points indicated in the text by
'''==> Complete data sheet'''
</p>

<h5>Report</h5>
<p>
Suggested guide/template for write up 10-15 pages
You should start by reading the discussions in e.g. Atkins and the handout itself.
Marks are given for correct results, but also for demonstrating that you understand how the results are obtained, and what they mean. A few carefully constructed diagrams with clear explanations are more likely to demonstrate this. Marks will be lost for unthinking copy/paste.
</p>

<h6>Exercise 1:</h6>
<p>
'''Theory''' <br/>
Briefly discuss the concepts of symmetry elements and symmetry operations, groups of symmetry operations (point groups) in your own words.
</p>

<p>
'''Results and Discussion''' <br/>
For the six molecules being tested, give the point group, flowchart assignment, and additional symmetry elements (from the data sheet). All symmetry elements should be illustrated for each molecule.
Comment on the symmetry elements that are present in both cyclohexane (demo) and chlorocubane, and those that are only present in one or other of the two molecules.
</p>
<h6>Exercise 2:</h6>

<p>
'''Theory''' <br/>
Briefly discuss the concept of an irreducible representation of a point group.
</p>

<p>
'''Results and Discussion''' <br/>
Tabulate and summarise the results you collected in completing the data sheet for each of the four examples.
What is a degenerate irreducible representation, and what symmetry element(s) must be present for these to occur? Discuss this, referring to the results you obtained for <math>BFBr_2</math> and <math>BCl_3</math>.
For <math>CrCO_6</math>, discuss the relationship between the correlation diagram, the orbitals you obtained (data sheet / diagrams) and their bonding characteristics. E.g. the <math>T_{2g}</math> orbitals are nonbonding and located mainly on the central metal atom.
</p>

<h6>Conclusions</h6>

<h3>[[Exercise 1 : Point groups - symmetry elements and operations]]</h3>

<h3>[[Exercise 2 : Symmetry properties of molecular vibrations and molecular orbitals]]</h3>

<h2>Molecular Symmetry : introduction to the lab, objectives</h2>

download .doc possible with image and media --- > the same at the end, a file that they can rewrite on... (forget the .htm extension)

[[Image:DATA SHEET_example_test1.doc]]
[[Media:DATA SHEET_example_test1.doc]]

<h3>Introduction</h3>

GFP : Green Fluorescent Protein

2009-08-10T16:25:44Z

Alasoro:

File:Gfp water charges amino acid.gjf

2009-08-10T16:25:09Z

Alasoro:

File:Gfp water charges.gjf

2009-08-10T16:24:53Z

Alasoro:

File:Gfp nowater charges.gjf

2009-08-10T16:24:30Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-10T16:24:09Z

Alasoro:

GFP : Green Fluorescent Protein

2009-08-10T16:20:20Z

Alasoro:

PYP : Photoactive yellow protein

2009-08-10T16:11:17Z

Alasoro:

<h4>PYP</h4>
[[Image:PYP_wiki_image.jpg]]

'''PYP, the Photoactive Yellow Protein''', is a small water-soluble protein extracted from the cytosol of the halophilic purple bacterium ''Halorhodospira halophila''. PYP is thought to mediate the phototactic response of the bacterium against blue light. Its chromophore is the deprotonated trans-p-hydroxycinnamic acid covalently linked, via a thioester bond, to the unique cysteine residue of the protein. Upon blue-light irradiation, PYP undergoes a photocycle. As for rhodopsins, the trans to cis isomerization of the chromophore was shown to be the first overall step of this photocycle.

[[Image:schema_pyp_pascal_changenet.gif]]

<h4>PDB : Protein Data Bank</h4>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

[[Image:cap_ecran_pyp_download.jpg]]

The PDB file has an extension which is '''.pdb''', and this extension can be directly open by '''GaussView''' (the short open is not possible, but Open GaussView and then click on File ---> Open and then select your .pdb file).

----
'''Break'''

If the coordinates of the hydrogens are not mentioned in the PDB file, GaussView will add the hydrogens automatically but so not necessary with the right angles and bond length. So be careful.
----

<h4>First calculation</h4>

Now you have the molecule that you want to study which is drawn so do exactly what you did usually so as to run a calculation using GaussView.

Just begin to run a single point energy calculation with AMBER.

You should get an error message with your first calculation (if not you are very lucky and so try to run the different calculations that you would like to run).

Normaly you should have so an error message which could be one of the different that are described in the [https://www.ch.ic.ac.uk/wiki/index.php/AMBER#Errors_fair AMBER part].

<h5>If it is a missing atom type :</h5>
have a look trough your input and find the atom which has no atom type, and give it one thanks to the different proposed by the [https://www.ch.ic.ac.uk/wiki/index.php/Some_data_useful_concerning_AMBER AMBER data base].

<h5>If it is missing MM parameters :</h5>
in this case operate very carefully as I mentioned below and do not be afraid to do it step by step.

First just consider the Bond length that are missing, identify the atoms corresponding in the molecule (with GaussView), and check for each of them that the atom type assigned by GaussView is the right one ! If the atom typ is wrong, so you have to change it in the input file. Once all the atom type verified (and changed for the wrong one) run again the calculation.

Now you should have some new missing parameters (or not!), if you have some new one have a look at the new atoms implicated in the new bond length missing and check if the atom types assigned are the right, if not change it...

And then run again the calculation...

When the Bond length missing do not involved some atom types not well assigned, then the Bond length and angles still missing have to be defined manually thanks to the [https://www.ch.ic.ac.uk/wiki/index.php/Some_data_useful_concerning_AMBER AMBER data base] (which is a cut of the AMBER paper).

----
'''Advice Break'''
In your protein you have a chromophore and using ONIOM enables you to run calculations with a high level of accuracy for this chromophore, so if some bond lengths or angles just involved atoms of the chromophore define them as 0.0

<pre>
HrmStr1 * * 0.0 0.0
HrmBnd1 * * * 0.0 0.0
</pre>

You can do this big approximation because/thanks to ONIOM theory : indeed an AMBER calculation will be run on the full protein and on the chromophore in order to deduce the contribution of the chromophore of the energy of the full protein (because concerning the chromophore the energy will be the one given by the high level of theory).

----

<h4>Example : PYP with 3PHY.pdb file</h4>

<h5>How to run the calculation ?</h5>

Download the .pdb file on the Protein Data Bank website : [http://www.rcsb.org/pdb/explore.do?structureId=3PHY 3phy:PHOTOACTIVE YELLOW PROTEIN, DARK STATE (UNBLEACHED), SOLUTION STRUCTURE, NMR, 26 STRUCTURES]

[[Media:3PHY.pdb]]

And try to run an AMBER calculation

<pre>
#p amber geom=connectivity
</pre>

So you should get this error message :
<pre>
Missing atomic parameters for atom 1914 IAtTyp= 20000000
Missing atomic parameters.
</pre>

Thus the atom 1914 is not define (atom type not define).

Indeed the O is not defined
<pre>
O- -9.64963601 -6.71154940 3.64049197
</pre>
Have a look at the environment : the O is an O type so write

<pre>
O-O- -9.64963601 -6.71154940 3.64049197
</pre>

Run again the calculation after this change.

An other error message should appear :

<pre>
Missing atomic parameters for atom 1923 IAtTyp= 20000000
Missing atomic parameters.
</pre>

Define this other atom and run again the calculation.

This time : a new error message, all the parameters below are missing, so have a look at the different atoms and atom types assigned by GaussView

<pre>
Include all MM classes
Bondstretch undefined between atoms 40 41 CM-N2
Bondstretch undefined between atoms 42 44 CM-N2
Bondstretch undefined between atoms 43 51 CA-H5
Bondstretch undefined between atoms 1003 1913 S-C
Bondstretch undefined between atoms 1610 1611 CM-N2
Bondstretch undefined between atoms 1612 1614 CM-N2
Bondstretch undefined between atoms 1613 1621 CA-H5
Bondstretch undefined between atoms 1780 1782 CM-N2
Bondstretch undefined between atoms 1915 1924 CM-HC
Bondstretch undefined between atoms 1916 1925 CM-HC
Angle bend undefined between atoms 36 39 40 CT-CT-CM
Angle bend undefined between atoms 39 40 41 CT-CM-N2
Angle bend undefined between atoms 40 41 43 CM-N2-CA
Angle bend undefined between atoms 40 41 49 CM-N2-H
Angle bend undefined between atoms 40 42 44 CM-CM-N2
Angle bend undefined between atoms 41 43 51 N2-CA-H5
Angle bend undefined between atoms 41 40 42 N2-CM-CM
Angle bend undefined between atoms 42 44 1930 CM-N2-H
Angle bend undefined between atoms 42 44 43 CM-N2-CA
Angle bend undefined between atoms 44 42 50 N2-CM-H4
Angle bend undefined between atoms 44 43 51 N2-CA-H5
Angle bend undefined between atoms 649 648 650 O2-C-OH
Angle bend undefined between atoms 1002 1003 1913 CT-S-C
Angle bend undefined between atoms 1003 1913 1914 S-C-O
Angle bend undefined between atoms 1003 1913 1915 S-C-CM
Angle bend undefined between atoms 1606 1609 1610 CT-CT-CM
Angle bend undefined between atoms 1609 1610 1611 CT-CM-N2
Angle bend undefined between atoms 1610 1611 1613 CM-N2-CA
Angle bend undefined between atoms 1610 1611 1619 CM-N2-H
Angle bend undefined between atoms 1610 1612 1614 CM-CM-N2
Angle bend undefined between atoms 1611 1613 1621 N2-CA-H5
Angle bend undefined between atoms 1611 1610 1612 N2-CM-CM
Angle bend undefined between atoms 1612 1614 1931 CM-N2-H
Angle bend undefined between atoms 1612 1614 1613 CM-N2-CA
Angle bend undefined between atoms 1614 1612 1620 N2-CM-H4
Angle bend undefined between atoms 1614 1613 1621 N2-CA-H5
Angle bend undefined between atoms 1775 1778 1779 CT-CT-CM
Angle bend undefined between atoms 1779 1780 1782 CM-CM-N2
Angle bend undefined between atoms 1779 1781 1784 CM-CM-CM
Angle bend undefined between atoms 1780 1782 1783 CM-N2-CA
Angle bend undefined between atoms 1780 1782 1793 CM-N2-H
Angle bend undefined between atoms 1780 1779 1781 CM-CM-CM
Angle bend undefined between atoms 1781 1783 1785 CM-CA-CA
Angle bend undefined between atoms 1782 1783 1785 N2-CA-CA
Angle bend undefined between atoms 1782 1780 1792 N2-CM-H4
Angle bend undefined between atoms 1784 1786 1787 CM-CA-CA
Angle bend undefined between atoms 1784 1786 1796 CM-CA-HA
Angle bend undefined between atoms 1913 1915 1924 C-CM-HC
Angle bend undefined between atoms 1915 1916 1925 CM-CM-HC
Angle bend undefined between atoms 1916 1917 1918 CM-CA-CA
Angle bend undefined between atoms 1916 1917 1922 CM-CA-CA
Angle bend undefined between atoms 1916 1915 1924 CM-CM-HC
Angle bend undefined between atoms 1917 1916 1925 CA-CM-HC
Angle bend undefined between atoms 1919 1920 1923 CA-C-O
Angle bend undefined between atoms 1921 1920 1923 CA-C-O
MM function not complete
</pre>

Many atoms are not well defined by GaussView, so change them.

Finally you should have change the following atoms

<pre>
< 40 C-CM
< 41 N-N2
< 42 C-CM
< 43 C-CA
< 44 N-N2
< 649 O-O2
< 650 O-OH
< 1610 C-CM
< 1611 N-N2
< 1612 C-CM
< 1613 C-CA
< 1614 N-N2
< 1779 C-CM
< 1780 C-CM
< 1781 C-CM
< 1782 N-N2
< 1783 C-CA
< 1784 C-CM
</pre>

By the following one

<pre>
> 40 C-CC
> 41 N-NA
> 42 C-CV
> 43 C-CR
> 44 N-NB
> 649 O-OH
> 650 O-O
> 1610 C-CC
> 1611 N-NA
> 1612 C-CV
> 1613 C-CR
> 1614 N-NB
> 1779 C-C*
> 1780 C-CW
> 1781 C-CB
> 1782 N-NA
> 1783 C-CN
> 1784 C-CA
</pre>

Now run again the calculation and you have to define manually the other bond lengths and angles missing.

So finally you should have at the end of your input the following MM parameters

<pre>
HrmStr1 S C 227.0 1.810
HrmStr1 NB H 434.0 1.010
HrmStr1 * * 0.0 0.0
HrmBnd1 CV NB H 30.0 120.0
HrmBnd1 CR NB H 30.0 120.0
HrmBnd1 O2 C OH 80.0 125.0
HrmBnd1 CT S C 62.0 100.0
HrmBnd1 S C O 80.0 124.0
HrmBnd1 S C CM 80.0 125.30
HrmBnd1 * * * 0.0 0.0
</pre>

<h5>Input, output fair</h5>

One atom has no atom type : [[Media:wiki_test_3phy_1.gjf]]
[[Media:wiki_test_3phy_1.log]]

One other atom has no atom type : [[Media:wiki_test_3phy_2.gjf]]
[[Media:wiki_test_3phy_2.log]]

Now we know that we have some MM parameters missing : begin to check the atoms types concerned [[Media:wiki_test_3phy_3.gjf]]
[[Media:wiki_test_3phy_3.log]]

First changes operated concerning atoms num 40 41 42 43 44 1610 1611 1612 1613 1614 1780 1782, and then run wiki_test_3phy_4

[[Media:wiki_test_3phy_4.gjf]]
[[Media:wiki_test_3phy_4.log]]

Change atom type of atom 1779 and then run wiki_test_3phy_5

[[Media:wiki_test_3phy_5.gjf]]
[[Media:wiki_test_3phy_5.log]]

Change atom type of atom 1781 and run wiki_test_3phy_6

[[Media:wiki_test_3phy_6.gjf]]
[[Media:wiki_test_3phy_6.log]]

Change atom type of atom 1784 and then run wiki_test_3phy_7

[[Media:wiki_test_3phy_7.gjf]]
[[Media:wiki_test_3phy_7.log]]

So this last should be ok

[[Media:wiki_test_3phy_8.gjf]]
[[Media:wiki_test_3phy_8.log]]

<h5>Atoms charges</h5>
As you can see in order to run an AMBER calculations the charges of all the atoms have to be specified. GaussView is able to specify many charges (the charges correspondent to different amino acid mentioned in the AMBER paper).

<pre>
N-N3-0.159200 0 22.53936399 -2.55554940 12.12649197 L
C-CT-0.022100 0 21.34836399 -2.49054940 11.23349197 L
C-C-0.612300 0 20.19836399 -3.28254940 11.85849197 L
O-O--0.571300 0 20.40836399 -4.18754940 12.64249197 L
C-CT-0.086500 0 21.70036399 -3.08954940 9.87049197 L
C-CT-0.033400 0 21.58536399 -2.01054940 8.79149197 L
S-S--0.277400 0 19.86636399 -1.89254940 8.23649197 L
C-CT--0.034100 0 20.13836399 -0.68754940 6.91449197 L
H-H-0.198400 0 22.76336399 -3.55054940 12.33349197 L
H-H-0.198400 0 23.35136399 -2.10754940 11.65549197 L
H-H-0.198400 0 22.33636399 -2.05454940 13.01449197 L
H-HP-0.111600 0 21.04836399 -1.46054940 11.10849197 L
H-HC-0.012500 0 22.71136399 -3.46954940 9.89449197 L
H-HC-0.012500 0 21.01736399 -3.89554940 9.64449197 L
H-H1-0.029200 0 21.89836399 -1.05954940 9.19849197 L
H-H1-0.029200 0 22.21736399 -2.26954940 7.95549197 L
H-H1-0.059700 0 20.92536399 -1.04054940 6.26349197 L
H-H1-0.059700 0 19.22536399 -0.56354940 6.34849197 L
H-H1-0.059700 0 20.42636399 0.25845060 7.34249197 L
</pre>

But some charges could be missing

<pre>
O-O--0.567900 0 3.95036399 -10.02254940 -0.92050803 L
C-CT--0.182500 0 6.36136399 -7.53054940 -0.90950803 L
H-H-0.271900 0 6.06136399 -7.38354940 -3.47850803 L
H-H1-0.082300 0 6.29536399 -9.54254940 -1.64450803 L
H-HC-0.060300 0 7.17236399 -7.08454940 -1.46550803 L
H-HC-0.060300 0 5.62336399 -6.77654940 -0.68050803 L
H-HC-0.060300 0 6.74436399 -7.95054940 0.00949197 L
N-N- 0 3.48336399 -7.89454940 -1.15250803 L
C-CT- 0 2.08936399 -8.10854940 -0.66650803 L
C-C- 0 1.35636399 -9.01354940 -1.64550803 L
O-O- 0 0.67236399 -9.94054940 -1.26750803 L
C-CT- 0 1.35836399 -6.77154940 -0.58450803 L
C-CT 0 1.44936399 -6.03754940 -1.92550803 L
C-C- 0 0.98036399 -4.59254940 -1.75050803 L
O-O2- 0 1.76836399 -3.70154940 -2.02150803 L
O-OH- 0 -0.30363601 -4.38854940 -1.29150803 L
H-H- 0 3.78536399 -6.99854940 -1.40350803 L
</pre>

That can correspond to some specific amino acid which are not well defined in the AMBER data base, or it can correspond to some unexpected sequences of atoms (in the case of a chromophore for example because it is not a specific amino acid, GaussView is not able to assign some charges).

'''If the charges missing correspond to an amino acid not well defined'''

In this case, have a look at the geometry and at the sequence of atoms, define the amino acid correspond and find in the AMBER paper the charges correspond to this case.

'''If you can not find a similar sequene in the AMBER paper'''
Most of the time it is the chromophore that is not defined (unexpected sequence of atoms).

I do not have an answer for the moment...
It seems that
[http://www.gaussian.com/g_tech/g_ur/k_mm.htm Amber=UnCharged] could be a good keyword, but I did not succeed running the calculation...

We'll see......

<h4>Files</h4>

I put different files here if you want to run some test calculations.

PYP with ONIOM, 2 layers (chromophore + surroundings) : [[Media:oniom_3phy_pyp_ready.gjf]]

<h4>Optimisation</h4>

Now that you have the complete input file, run an optimisation on the structure:

<pre>
#p opt=quadmac oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

Opt=Quadmac is mainly used for problem convergence cases, which does a quadratic step in the coordinates of all the atoms. Such an optimization can use either an updated approximate Hessian for the internal coordinates or an analytically computed Hessian. It computes products of the low-level (MM) Hessian with vectors as needed.

You must specify a ub3lyp method, the default rb3lyp method is not compatible with the system.

This job will terminate without completion:

<pre>
Error termination request processed by link 9999.
Error termination via Lnk1e in /home/gaussian-devel/gaussiandvh01_pgi_725/gdv/l9999.exe at Mon Aug 3 22:16:14 2009.
</pre>

On inspection of the error message and of the optimisation route in GaussView, it is apparent that the job ran out of steps and was fluctuating between two points. To fix this, copy the last geometry into a new input file and specify a smaller step size:

<pre>
#p opt=(quadmac,maxstep=10) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

This job completes normally.

<h5>Input and output fair</h5>
<p>
[[Media:3PHYopt1.gjf]]<br/>
Last geometry: [[Media: ERROR.gjf]]<br/>
[[Media:3PHYopt2.gjf]]<br/>
Final geometry: [[Media:DONE.log]]<br/>

<h4>Frequency</h4>

<h5>Freq, Freq=HPModes</h5>

Check that the optimised structure is really a minimum by running a joint optimisation and frequency calculation:

<pre>
#p opt=(quadmac,maxstep=10) freq oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

The geometry modifies slightly: [[Media:geometry.log]]<br/>

There are no imaginary frequencies, therefore this structure is a minimum (as a tip for ease in .log file navigation, use X11 to open the .log files with the "nedit" command).

<pre>
Low frequencies --- -0.3097 -0.1667 -0.0006 0.0004 0.0006 0.1446
Low frequencies --- 4.5860 5.7688 7.1004
Diagonal vibrational polarizability:
5570.9936676 5790.9109142 7396.2535226
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1 2 3
A A A
Frequencies -- 4.5858 5.7681 7.1004
Red. masses -- 5.6257 7.1713 6.7230
Frc consts -- 0.0001 0.0001 0.0002
%ModelSys -- 0.8136 5.6385 1.5914
%RealSys -- 99.1864 94.3615 98.4086
IR Inten -- 0.1580 0.7010 0.7201
Atom AN X Y Z X Y Z X Y Z
1 7 0.02 0.01 0.03 0.00 0.00 -0.01 0.02 0.01 0.02
2 6 0.02 0.01 0.03 0.00 0.01 -0.02 0.02 0.00 0.04
3 6 0.02 0.01 0.03 0.00 0.01 -0.01 0.02 0.00 0.04
4 8 0.03 0.01 0.04 0.00 0.00 -0.02 0.02 0.01 0.05
5 6 0.02 0.01 0.05 0.00 0.00 -0.02 0.02 0.01 0.05
</pre>

Now copy the same geometry into another input file (to avoid possible corruption with .chk) and run a freq=HPmodes calculation. This keyword gives the vibrational frequency eigenvectors to 5 figures (high precision):

<pre>
Low frequencies --- -0.3101 -0.1651 -0.0004 0.0001 0.0006 0.1435
Low frequencies --- 4.5908 5.7674 7.1007
Diagonal vibrational polarizability:
5570.6337105 5790.9612763 7395.8674830
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1 2 3 4 5
A A A A A
Frequencies --- 4.5906 5.7668 7.1006 7.3648 7.6765
Reduced masses --- 5.6278 7.1696 6.7235 7.0975 7.2004
Force constants --- 0.0001 0.0001 0.0002 0.0002 0.0002
Percent ModelSys --- 0.8203 5.6428 1.5962 0.1342 2.1399
Percent RealSys --- 99.1797 94.3572 98.4038 99.8658 97.8601
IR Intensities --- 0.1579 0.7002 0.7215 0.1148 0.4400
Coord Atom Element:
1 1 7 0.02278 -0.00267 0.02068 0.01019 0.03594
2 1 7 0.01230 0.00471 0.00633 0.05492 -0.00137
3 1 7 0.02519 -0.01023 0.02441 0.07149 -0.00477
</pre>

Note that both of these .log files are very large (in particular the second one which is 915mb with over 12.7 million lines). Consequently your system may crash if too many processes are running at once.

<h5>Freq=ModelModes</h5>

Run a freq=modelmodes calculation on the same geometry.

<pre>
#p freq=(modelmodes) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

This keyword displays modes associated with the small model system in this ONIOM calculation:

<pre>
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1180 1301 1592
A A A
Frequencies -- 347.2729 400.0784 528.7142
</pre>

Only 31 modes instead of 45? (Would it be possible that Gaussian erase some modes because the file was too big ??)

<h5>Freq=(HPModes,ModelModes)</h5>

Finally run a freq=(HPmodes,modelmodes) calculation:

<pre>
#p freq=(hpmodes,modelmodes) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

In comparison, the last two files are much smaller than the first two and therefore easier to work with.

If you compute the calculation with '''Freq=(HPModes,ModelModes) keyword''' you directly get the informations that you need :

* First, you know if you have some imaginary frequencies (no mention of the imaginary frequency if there is no).
<pre>
Full mass-weighted force constant matrix:
Low frequencies --- -727.1752 -603.3920 -441.6366 -348.9122 -336.0877 -330.3771
Low frequencies --- -318.1434 -313.7065 -289.1518
****** 362 imaginary frequencies (negative Signs) ******
Diagonal vibrational polarizability:
52770.5393600 27276.1819608 39738.0499572
NorSel: MapVib= 2181 2332 2360 2364 2521 2847 3397 3653 3874 4190
NorSel: MapVib= 4714 4764 4822 4839 5543 5580 5585 5736 5765 5766
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
2181 2332 2360
A A A
Frequencies -- 781.7748 858.7045 871.9542
Red. masses -- 4.5651 1.7393 1.9629
</pre>

* Second, the precision will be better, because the '''HPModes keyword''' enable you to get 5 figures for the vibrational frequency eigenvectors, which could be useful so as to run some other calculation to et some overlap.

<h5>Input fair</h5>

[[Media:3PHYopt1_1freq.gjf]]<br/>
[[Media:3PHYopt1_1freq2.gjf]]<br/>
[[Media:3PHYopt1_1freq3.gjf]]<br/>
[[Media:3PHYopt1_1freq4.gjf]]<br/>
[[Media:3PHYopt1_1freqMM.gjf]]<br/>

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

PYP : Photoactive yellow protein

2009-08-10T16:10:28Z

Alasoro:

<h4>PYP</h4>
[[Image:PYP_wiki_image.jpg]]

'''PYP, the Photoactive Yellow Protein''', is a small water-soluble protein extracted from the cytosol of the halophilic purple bacterium ''Halorhodospira halophila''. PYP is thought to mediate the phototactic response of the bacterium against blue light. Its chromophore is the deprotonated trans-p-hydroxycinnamic acid covalently linked, via a thioester bond, to the unique cysteine residue of the protein. Upon blue-light irradiation, PYP undergoes a photocycle. As for rhodopsins, the trans to cis isomerization of the chromophore was shown to be the first overall step of this photocycle.

[[Image:schema_pyp_pascal_changenet.gif]]

<h4>PDB : Protein Data Bank</h4>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

[[Image:cap_ecran_pyp_download.jpg]]

The PDB file has an extension which is '''.pdb''', and this extension can be directly open by '''GaussView''' (the short open is not possible, but Open GaussView and then click on File ---> Open and then select your .pdb file).

----
'''Break'''

If the coordinates of the hydrogens are not mentioned in the PDB file, GaussView will add the hydrogens automatically but so not necessary with the right angles and bond length. So be careful.
----

<h4>First calculation</h4>

Now you have the molecule that you want to study which is drawn so do exactly what you did usually so as to run a calculation using GaussView.

Just begin to run a single point energy calculation with AMBER.

You should get an error message with your first calculation (if not you are very lucky and so try to run the different calculations that you would like to run).

Normaly you should have so an error message which could be one of the different that are described in the [https://www.ch.ic.ac.uk/wiki/index.php/AMBER#Errors_fair AMBER part].

<h5>If it is a missing atom type :</h5>
have a look trough your input and find the atom which has no atom type, and give it one thanks to the different proposed by the [https://www.ch.ic.ac.uk/wiki/index.php/Some_data_useful_concerning_AMBER AMBER data base].

<h5>If it is missing MM parameters :</h5>
in this case operate very carefully as I mentioned below and do not be afraid to do it step by step.

First just consider the Bond length that are missing, identify the atoms corresponding in the molecule (with GaussView), and check for each of them that the atom type assigned by GaussView is the right one ! If the atom typ is wrong, so you have to change it in the input file. Once all the atom type verified (and changed for the wrong one) run again the calculation.

Now you should have some new missing parameters (or not!), if you have some new one have a look at the new atoms implicated in the new bond length missing and check if the atom types assigned are the right, if not change it...

And then run again the calculation...

When the Bond length missing do not involved some atom types not well assigned, then the Bond length and angles still missing have to be defined manually thanks to the [https://www.ch.ic.ac.uk/wiki/index.php/Some_data_useful_concerning_AMBER AMBER data base] (which is a cut of the AMBER paper).

----
'''Advice Break'''
In your protein you have a chromophore and using ONIOM enables you to run calculations with a high level of accuracy for this chromophore, so if some bond lengths or angles just involved atoms of the chromophore define them as 0.0

<pre>
HrmStr1 * * 0.0 0.0
HrmBnd1 * * * 0.0 0.0
</pre>

You can do this big approximation because/thanks to ONIOM theory : indeed an AMBER calculation will be run on the full protein and on the chromophore in order to deduce the contribution of the chromophore of the energy of the full protein (because concerning the chromophore the energy will be the one given by the high level of theory).

----

<h4>Example : PYP with 3PHY.pdb file</h4>

<h5>How to run the calculation ?</h5>

Download the .pdb file on the Protein Data Bank website : [http://www.rcsb.org/pdb/explore.do?structureId=3PHY 3phy:PHOTOACTIVE YELLOW PROTEIN, DARK STATE (UNBLEACHED), SOLUTION STRUCTURE, NMR, 26 STRUCTURES]

[[Media:3PHY.pdb]]

And try to run an AMBER calculation

<pre>
#p amber geom=connectivity
</pre>

So you should get this error message :
<pre>
Missing atomic parameters for atom 1914 IAtTyp= 20000000
Missing atomic parameters.
</pre>

Thus the atom 1914 is not define (atom type not define).

Indeed the O is not defined
<pre>
O- -9.64963601 -6.71154940 3.64049197
</pre>
Have a look at the environment : the O is an O type so write

<pre>
O-O- -9.64963601 -6.71154940 3.64049197
</pre>

Run again the calculation after this change.

An other error message should appear :

<pre>
Missing atomic parameters for atom 1923 IAtTyp= 20000000
Missing atomic parameters.
</pre>

Define this other atom and run again the calculation.

This time : a new error message, all the parameters below are missing, so have a look at the different atoms and atom types assigned by GaussView

<pre>
Include all MM classes
Bondstretch undefined between atoms 40 41 CM-N2
Bondstretch undefined between atoms 42 44 CM-N2
Bondstretch undefined between atoms 43 51 CA-H5
Bondstretch undefined between atoms 1003 1913 S-C
Bondstretch undefined between atoms 1610 1611 CM-N2
Bondstretch undefined between atoms 1612 1614 CM-N2
Bondstretch undefined between atoms 1613 1621 CA-H5
Bondstretch undefined between atoms 1780 1782 CM-N2
Bondstretch undefined between atoms 1915 1924 CM-HC
Bondstretch undefined between atoms 1916 1925 CM-HC
Angle bend undefined between atoms 36 39 40 CT-CT-CM
Angle bend undefined between atoms 39 40 41 CT-CM-N2
Angle bend undefined between atoms 40 41 43 CM-N2-CA
Angle bend undefined between atoms 40 41 49 CM-N2-H
Angle bend undefined between atoms 40 42 44 CM-CM-N2
Angle bend undefined between atoms 41 43 51 N2-CA-H5
Angle bend undefined between atoms 41 40 42 N2-CM-CM
Angle bend undefined between atoms 42 44 1930 CM-N2-H
Angle bend undefined between atoms 42 44 43 CM-N2-CA
Angle bend undefined between atoms 44 42 50 N2-CM-H4
Angle bend undefined between atoms 44 43 51 N2-CA-H5
Angle bend undefined between atoms 649 648 650 O2-C-OH
Angle bend undefined between atoms 1002 1003 1913 CT-S-C
Angle bend undefined between atoms 1003 1913 1914 S-C-O
Angle bend undefined between atoms 1003 1913 1915 S-C-CM
Angle bend undefined between atoms 1606 1609 1610 CT-CT-CM
Angle bend undefined between atoms 1609 1610 1611 CT-CM-N2
Angle bend undefined between atoms 1610 1611 1613 CM-N2-CA
Angle bend undefined between atoms 1610 1611 1619 CM-N2-H
Angle bend undefined between atoms 1610 1612 1614 CM-CM-N2
Angle bend undefined between atoms 1611 1613 1621 N2-CA-H5
Angle bend undefined between atoms 1611 1610 1612 N2-CM-CM
Angle bend undefined between atoms 1612 1614 1931 CM-N2-H
Angle bend undefined between atoms 1612 1614 1613 CM-N2-CA
Angle bend undefined between atoms 1614 1612 1620 N2-CM-H4
Angle bend undefined between atoms 1614 1613 1621 N2-CA-H5
Angle bend undefined between atoms 1775 1778 1779 CT-CT-CM
Angle bend undefined between atoms 1779 1780 1782 CM-CM-N2
Angle bend undefined between atoms 1779 1781 1784 CM-CM-CM
Angle bend undefined between atoms 1780 1782 1783 CM-N2-CA
Angle bend undefined between atoms 1780 1782 1793 CM-N2-H
Angle bend undefined between atoms 1780 1779 1781 CM-CM-CM
Angle bend undefined between atoms 1781 1783 1785 CM-CA-CA
Angle bend undefined between atoms 1782 1783 1785 N2-CA-CA
Angle bend undefined between atoms 1782 1780 1792 N2-CM-H4
Angle bend undefined between atoms 1784 1786 1787 CM-CA-CA
Angle bend undefined between atoms 1784 1786 1796 CM-CA-HA
Angle bend undefined between atoms 1913 1915 1924 C-CM-HC
Angle bend undefined between atoms 1915 1916 1925 CM-CM-HC
Angle bend undefined between atoms 1916 1917 1918 CM-CA-CA
Angle bend undefined between atoms 1916 1917 1922 CM-CA-CA
Angle bend undefined between atoms 1916 1915 1924 CM-CM-HC
Angle bend undefined between atoms 1917 1916 1925 CA-CM-HC
Angle bend undefined between atoms 1919 1920 1923 CA-C-O
Angle bend undefined between atoms 1921 1920 1923 CA-C-O
MM function not complete
</pre>

Many atoms are not well defined by GaussView, so change them.

Finally you should have change the following atoms

<pre>
< 40 C-CM
< 41 N-N2
< 42 C-CM
< 43 C-CA
< 44 N-N2
< 649 O-O2
< 650 O-OH
< 1610 C-CM
< 1611 N-N2
< 1612 C-CM
< 1613 C-CA
< 1614 N-N2
< 1779 C-CM
< 1780 C-CM
< 1781 C-CM
< 1782 N-N2
< 1783 C-CA
< 1784 C-CM
</pre>

By the following one

<pre>
> 40 C-CC
> 41 N-NA
> 42 C-CV
> 43 C-CR
> 44 N-NB
> 649 O-OH
> 650 O-O
> 1610 C-CC
> 1611 N-NA
> 1612 C-CV
> 1613 C-CR
> 1614 N-NB
> 1779 C-C*
> 1780 C-CW
> 1781 C-CB
> 1782 N-NA
> 1783 C-CN
> 1784 C-CA
</pre>

Now run again the calculation and you have to define manually the other bond lengths and angles missing.

So finally you should have at the end of your input the following MM parameters

<pre>
HrmStr1 S C 227.0 1.810
HrmStr1 NB H 434.0 1.010
HrmStr1 * * 0.0 0.0
HrmBnd1 CV NB H 30.0 120.0
HrmBnd1 CR NB H 30.0 120.0
HrmBnd1 O2 C OH 80.0 125.0
HrmBnd1 CT S C 62.0 100.0
HrmBnd1 S C O 80.0 124.0
HrmBnd1 S C CM 80.0 125.30
HrmBnd1 * * * 0.0 0.0
</pre>

<h5>Input, output fair</h5>

One atom has no atom type : [[Media:wiki_test_3phy_1.gjf]]
[[Media:wiki_test_3phy_1.log]]

One other atom has no atom type : [[Media:wiki_test_3phy_2.gjf]]
[[Media:wiki_test_3phy_2.log]]

Now we know that we have some MM parameters missing : begin to check the atoms types concerned [[Media:wiki_test_3phy_3.gjf]]
[[Media:wiki_test_3phy_3.log]]

First changes operated concerning atoms num 40 41 42 43 44 1610 1611 1612 1613 1614 1780 1782, and then run wiki_test_3phy_4

[[Media:wiki_test_3phy_4.gjf]]
[[Media:wiki_test_3phy_4.log]]

Change atom type of atom 1779 and then run wiki_test_3phy_5

[[Media:wiki_test_3phy_5.gjf]]
[[Media:wiki_test_3phy_5.log]]

Change atom type of atom 1781 and run wiki_test_3phy_6

[[Media:wiki_test_3phy_6.gjf]]
[[Media:wiki_test_3phy_6.log]]

Change atom type of atom 1784 and then run wiki_test_3phy_7

[[Media:wiki_test_3phy_7.gjf]]
[[Media:wiki_test_3phy_7.log]]

So this last should be ok

[[Media:wiki_test_3phy_8.gjf]]
[[Media:wiki_test_3phy_8.log]]

<h5>Atoms charges</h5>
As you can see in order to run an AMBER calculations the charges of all the atoms have to be specified. GaussView is able to specify many charges (the charges correspondent to different amino acid mentioned in the AMBER paper).

<pre>
N-N3-0.159200 0 22.53936399 -2.55554940 12.12649197 L
C-CT-0.022100 0 21.34836399 -2.49054940 11.23349197 L
C-C-0.612300 0 20.19836399 -3.28254940 11.85849197 L
O-O--0.571300 0 20.40836399 -4.18754940 12.64249197 L
C-CT-0.086500 0 21.70036399 -3.08954940 9.87049197 L
C-CT-0.033400 0 21.58536399 -2.01054940 8.79149197 L
S-S--0.277400 0 19.86636399 -1.89254940 8.23649197 L
C-CT--0.034100 0 20.13836399 -0.68754940 6.91449197 L
H-H-0.198400 0 22.76336399 -3.55054940 12.33349197 L
H-H-0.198400 0 23.35136399 -2.10754940 11.65549197 L
H-H-0.198400 0 22.33636399 -2.05454940 13.01449197 L
H-HP-0.111600 0 21.04836399 -1.46054940 11.10849197 L
H-HC-0.012500 0 22.71136399 -3.46954940 9.89449197 L
H-HC-0.012500 0 21.01736399 -3.89554940 9.64449197 L
H-H1-0.029200 0 21.89836399 -1.05954940 9.19849197 L
H-H1-0.029200 0 22.21736399 -2.26954940 7.95549197 L
H-H1-0.059700 0 20.92536399 -1.04054940 6.26349197 L
H-H1-0.059700 0 19.22536399 -0.56354940 6.34849197 L
H-H1-0.059700 0 20.42636399 0.25845060 7.34249197 L
</pre>

But some charges could be missing

<pre>
O-O--0.567900 0 3.95036399 -10.02254940 -0.92050803 L
C-CT--0.182500 0 6.36136399 -7.53054940 -0.90950803 L
H-H-0.271900 0 6.06136399 -7.38354940 -3.47850803 L
H-H1-0.082300 0 6.29536399 -9.54254940 -1.64450803 L
H-HC-0.060300 0 7.17236399 -7.08454940 -1.46550803 L
H-HC-0.060300 0 5.62336399 -6.77654940 -0.68050803 L
H-HC-0.060300 0 6.74436399 -7.95054940 0.00949197 L
N-N- 0 3.48336399 -7.89454940 -1.15250803 L
C-CT- 0 2.08936399 -8.10854940 -0.66650803 L
C-C- 0 1.35636399 -9.01354940 -1.64550803 L
O-O- 0 0.67236399 -9.94054940 -1.26750803 L
C-CT- 0 1.35836399 -6.77154940 -0.58450803 L
C-CT 0 1.44936399 -6.03754940 -1.92550803 L
C-C- 0 0.98036399 -4.59254940 -1.75050803 L
O-O2- 0 1.76836399 -3.70154940 -2.02150803 L
O-OH- 0 -0.30363601 -4.38854940 -1.29150803 L
H-H- 0 3.78536399 -6.99854940 -1.40350803 L
</pre>

That can correspond to some specific amino acid which are not well defined in the AMBER data base, or it can correspond to some unexpected sequences of atoms (in the case of a chromophore for example because it is not a specific amino acid, GaussView is not able to assign some charges).

'''If the charges missing correspond to an amino acid not well defined'''

In this case, have a look at the geometry and at the sequence of atoms, define the amino acid correspond and find in the AMBER paper the charges correspond to this case.

'''If you can not find a similar sequene in the AMBER paper'''
Most of the time it is the chromophore that is not defined (unexpected sequence of atoms).

I do not have an answer for the moment...
It seems that
[http://www.gaussian.com/g_tech/g_ur/k_mm.htm Amber=UnCharged] could be a good keyword, but I did not succeed running the calculation...

We'll see......

<h4>Files</h4>

I put different files here if you want to run some test calculations.

PYP with ONIOM, 2 layers (chromophore + surroundings) : [[Media:oniom_3phy_pyp_ready.gjf]]

<h4>Optimisation</h4>

Now that you have the complete input file, run an optimisation on the structure:

<pre>
#p opt=quadmac oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

Opt=Quadmac is mainly used for problem convergence cases, which does a quadratic step in the coordinates of all the atoms. Such an optimization can use either an updated approximate Hessian for the internal coordinates or an analytically computed Hessian. It computes products of the low-level (MM) Hessian with vectors as needed.

You must specify a ub3lyp method, the default rb3lyp method is not compatible with the system.

This job will terminate without completion:

<pre>
Error termination request processed by link 9999.
Error termination via Lnk1e in /home/gaussian-devel/gaussiandvh01_pgi_725/gdv/l9999.exe at Mon Aug 3 22:16:14 2009.
</pre>

On inspection of the error message and of the optimisation route in GaussView, it is apparent that the job ran out of steps and was fluctuating between two points. To fix this, copy the last geometry into a new input file and specify a smaller step size:

<pre>
#p opt=(quadmac,maxstep=10) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

This job completes normally.

<h5>Input and output fair</h5>
<p>
[[Media:3PHYopt1.gjf]]<br/>
Last geometry: [[Media: ERROR.gjf]]<br/>
[[Media:3PHYopt2.gjf]]<br/>
Final geometry: [[Media:DONE.log]]<br/>

<h4>Frequency</h4>

<h5>Freq, Freq=HPModes</h5>

Check that the optimised structure is really a minimum by running a joint optimisation and frequency calculation:

<pre>
#p opt=(quadmac,maxstep=10) freq oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

The geometry modifies slightly: [[Media:geometry.log]]<br/>

There are no imaginary frequencies, therefore this structure is a minimum (as a tip for ease in .log file navigation, use X11 to open the .log files with the "nedit" command).

<pre>
Low frequencies --- -0.3097 -0.1667 -0.0006 0.0004 0.0006 0.1446
Low frequencies --- 4.5860 5.7688 7.1004
Diagonal vibrational polarizability:
5570.9936676 5790.9109142 7396.2535226
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1 2 3
A A A
Frequencies -- 4.5858 5.7681 7.1004
Red. masses -- 5.6257 7.1713 6.7230
Frc consts -- 0.0001 0.0001 0.0002
%ModelSys -- 0.8136 5.6385 1.5914
%RealSys -- 99.1864 94.3615 98.4086
IR Inten -- 0.1580 0.7010 0.7201
Atom AN X Y Z X Y Z X Y Z
1 7 0.02 0.01 0.03 0.00 0.00 -0.01 0.02 0.01 0.02
2 6 0.02 0.01 0.03 0.00 0.01 -0.02 0.02 0.00 0.04
3 6 0.02 0.01 0.03 0.00 0.01 -0.01 0.02 0.00 0.04
4 8 0.03 0.01 0.04 0.00 0.00 -0.02 0.02 0.01 0.05
5 6 0.02 0.01 0.05 0.00 0.00 -0.02 0.02 0.01 0.05
</pre>

Now copy the same geometry into another input file (to avoid possible corruption with .chk) and run a freq=HPmodes calculation. This keyword gives the vibrational frequency eigenvectors to 5 figures (high precision):

<pre>
Low frequencies --- -0.3101 -0.1651 -0.0004 0.0001 0.0006 0.1435
Low frequencies --- 4.5908 5.7674 7.1007
Diagonal vibrational polarizability:
5570.6337105 5790.9612763 7395.8674830
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1 2 3 4 5
A A A A A
Frequencies --- 4.5906 5.7668 7.1006 7.3648 7.6765
Reduced masses --- 5.6278 7.1696 6.7235 7.0975 7.2004
Force constants --- 0.0001 0.0001 0.0002 0.0002 0.0002
Percent ModelSys --- 0.8203 5.6428 1.5962 0.1342 2.1399
Percent RealSys --- 99.1797 94.3572 98.4038 99.8658 97.8601
IR Intensities --- 0.1579 0.7002 0.7215 0.1148 0.4400
Coord Atom Element:
1 1 7 0.02278 -0.00267 0.02068 0.01019 0.03594
2 1 7 0.01230 0.00471 0.00633 0.05492 -0.00137
3 1 7 0.02519 -0.01023 0.02441 0.07149 -0.00477
</pre>

Note that both of these .log files are very large (in particular the second one which is 915mb with over 12.7 million lines). Consequently your system may crash if too many processes are running at once.

<h5>Freq=ModelModes</h5>

Run a freq=modelmodes calculation on the same geometry.

<pre>
#p freq=(modelmodes) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

This keyword displays modes associated with the small model system in this ONIOM calculation:

<pre>
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1180 1301 1592
A A A
Frequencies -- 347.2729 400.0784 528.7142
</pre>

Only 31 modes instead of 45?

<h5>Freq=(HPModes,ModelModes)</h5>

Finally run a freq=(HPmodes,modelmodes) calculation:

<pre>
#p freq=(hpmodes,modelmodes) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

In comparison, the last two files are much smaller than the first two and therefore easier to work with.

If you compute the calculation with '''Freq=(HPModes,ModelModes) keyword''' you directly get the informations that you need :

* First, you know if you have some imaginary frequencies (no mention of the imaginary frequency if there is no).
<pre>
Full mass-weighted force constant matrix:
Low frequencies --- -727.1752 -603.3920 -441.6366 -348.9122 -336.0877 -330.3771
Low frequencies --- -318.1434 -313.7065 -289.1518
****** 362 imaginary frequencies (negative Signs) ******
Diagonal vibrational polarizability:
52770.5393600 27276.1819608 39738.0499572
NorSel: MapVib= 2181 2332 2360 2364 2521 2847 3397 3653 3874 4190
NorSel: MapVib= 4714 4764 4822 4839 5543 5580 5585 5736 5765 5766
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
2181 2332 2360
A A A
Frequencies -- 781.7748 858.7045 871.9542
Red. masses -- 4.5651 1.7393 1.9629
</pre>

* Second, the precision will be better, because the '''HPModes keyword''' enable you to get 5 figures for the vibrational frequency eigenvectors, which could be useful so as to run some other calculation to et some overlap.

<h5>Input fair</h5>

[[Media:3PHYopt1_1freq.gjf]]<br/>
[[Media:3PHYopt1_1freq2.gjf]]<br/>
[[Media:3PHYopt1_1freq3.gjf]]<br/>
[[Media:3PHYopt1_1freq4.gjf]]<br/>
[[Media:3PHYopt1_1freqMM.gjf]]<br/>

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]

PYP : Photoactive yellow protein

2009-08-10T16:09:52Z

Alasoro:

<h4>PYP</h4>
[[Image:PYP_wiki_image.jpg]]

'''PYP, the Photoactive Yellow Protein''', is a small water-soluble protein extracted from the cytosol of the halophilic purple bacterium ''Halorhodospira halophila''. PYP is thought to mediate the phototactic response of the bacterium against blue light. Its chromophore is the deprotonated trans-p-hydroxycinnamic acid covalently linked, via a thioester bond, to the unique cysteine residue of the protein. Upon blue-light irradiation, PYP undergoes a photocycle. As for rhodopsins, the trans to cis isomerization of the chromophore was shown to be the first overall step of this photocycle.

[[Image:schema_pyp_pascal_changenet.gif]]

<h4>PDB : Protein Data Bank</h4>

In order to run some calculations about a protein, you need an input file ! You can find the protein already drawn in the [http://www.rcsb.org/pdb/home/home.do Protein Data Bank (PDB)] website. Just search your molecule in this data base which regroup most research already done about protein. Choose a file, and then click on '''Download files --- > PDB file (Text)'''.

[[Image:cap_ecran_pyp_download.jpg]]

The PDB file has an extension which is '''.pdb''', and this extension can be directly open by '''GaussView''' (the short open is not possible, but Open GaussView and then click on File ---> Open and then select your .pdb file).

----
'''Break'''

If the coordinates of the hydrogens are not mentioned in the PDB file, GaussView will add the hydrogens automatically but so not necessary with the right angles and bond length. So be careful.
----

<h4>First calculation</h4>

Now you have the molecule that you want to study which is drawn so do exactly what you did usually so as to run a calculation using GaussView.

Just begin to run a single point energy calculation with AMBER.

You should get an error message with your first calculation (if not you are very lucky and so try to run the different calculations that you would like to run).

Normaly you should have so an error message which could be one of the different that are described in the [https://www.ch.ic.ac.uk/wiki/index.php/AMBER#Errors_fair AMBER part].

<h5>If it is a missing atom type :</h5>
have a look trough your input and find the atom which has no atom type, and give it one thanks to the different proposed by the [https://www.ch.ic.ac.uk/wiki/index.php/Some_data_useful_concerning_AMBER AMBER data base].

<h5>If it is missing MM parameters :</h5>
in this case operate very carefully as I mentioned below and do not be afraid to do it step by step.

First just consider the Bond length that are missing, identify the atoms corresponding in the molecule (with GaussView), and check for each of them that the atom type assigned by GaussView is the right one ! If the atom typ is wrong, so you have to change it in the input file. Once all the atom type verified (and changed for the wrong one) run again the calculation.

Now you should have some new missing parameters (or not!), if you have some new one have a look at the new atoms implicated in the new bond length missing and check if the atom types assigned are the right, if not change it...

And then run again the calculation...

When the Bond length missing do not involved some atom types not well assigned, then the Bond length and angles still missing have to be defined manually thanks to the [https://www.ch.ic.ac.uk/wiki/index.php/Some_data_useful_concerning_AMBER AMBER data base] (which is a cut of the AMBER paper).

----
'''Advice Break'''
In your protein you have a chromophore and using ONIOM enables you to run calculations with a high level of accuracy for this chromophore, so if some bond lengths or angles just involved atoms of the chromophore define them as 0.0

<pre>
HrmStr1 * * 0.0 0.0
HrmBnd1 * * * 0.0 0.0
</pre>

You can do this big approximation because/thanks to ONIOM theory : indeed an AMBER calculation will be run on the full protein and on the chromophore in order to deduce the contribution of the chromophore of the energy of the full protein (because concerning the chromophore the energy will be the one given by the high level of theory).

----

<h4>Example : PYP with 3PHY.pdb file</h4>

<h5>How to run the calculation ?</h5>

Download the .pdb file on the Protein Data Bank website : [http://www.rcsb.org/pdb/explore.do?structureId=3PHY 3phy:PHOTOACTIVE YELLOW PROTEIN, DARK STATE (UNBLEACHED), SOLUTION STRUCTURE, NMR, 26 STRUCTURES]

[[Media:3PHY.pdb]]

And try to run an AMBER calculation

<pre>
#p amber geom=connectivity
</pre>

So you should get this error message :
<pre>
Missing atomic parameters for atom 1914 IAtTyp= 20000000
Missing atomic parameters.
</pre>

Thus the atom 1914 is not define (atom type not define).

Indeed the O is not defined
<pre>
O- -9.64963601 -6.71154940 3.64049197
</pre>
Have a look at the environment : the O is an O type so write

<pre>
O-O- -9.64963601 -6.71154940 3.64049197
</pre>

Run again the calculation after this change.

An other error message should appear :

<pre>
Missing atomic parameters for atom 1923 IAtTyp= 20000000
Missing atomic parameters.
</pre>

Define this other atom and run again the calculation.

This time : a new error message, all the parameters below are missing, so have a look at the different atoms and atom types assigned by GaussView

<pre>
Include all MM classes
Bondstretch undefined between atoms 40 41 CM-N2
Bondstretch undefined between atoms 42 44 CM-N2
Bondstretch undefined between atoms 43 51 CA-H5
Bondstretch undefined between atoms 1003 1913 S-C
Bondstretch undefined between atoms 1610 1611 CM-N2
Bondstretch undefined between atoms 1612 1614 CM-N2
Bondstretch undefined between atoms 1613 1621 CA-H5
Bondstretch undefined between atoms 1780 1782 CM-N2
Bondstretch undefined between atoms 1915 1924 CM-HC
Bondstretch undefined between atoms 1916 1925 CM-HC
Angle bend undefined between atoms 36 39 40 CT-CT-CM
Angle bend undefined between atoms 39 40 41 CT-CM-N2
Angle bend undefined between atoms 40 41 43 CM-N2-CA
Angle bend undefined between atoms 40 41 49 CM-N2-H
Angle bend undefined between atoms 40 42 44 CM-CM-N2
Angle bend undefined between atoms 41 43 51 N2-CA-H5
Angle bend undefined between atoms 41 40 42 N2-CM-CM
Angle bend undefined between atoms 42 44 1930 CM-N2-H
Angle bend undefined between atoms 42 44 43 CM-N2-CA
Angle bend undefined between atoms 44 42 50 N2-CM-H4
Angle bend undefined between atoms 44 43 51 N2-CA-H5
Angle bend undefined between atoms 649 648 650 O2-C-OH
Angle bend undefined between atoms 1002 1003 1913 CT-S-C
Angle bend undefined between atoms 1003 1913 1914 S-C-O
Angle bend undefined between atoms 1003 1913 1915 S-C-CM
Angle bend undefined between atoms 1606 1609 1610 CT-CT-CM
Angle bend undefined between atoms 1609 1610 1611 CT-CM-N2
Angle bend undefined between atoms 1610 1611 1613 CM-N2-CA
Angle bend undefined between atoms 1610 1611 1619 CM-N2-H
Angle bend undefined between atoms 1610 1612 1614 CM-CM-N2
Angle bend undefined between atoms 1611 1613 1621 N2-CA-H5
Angle bend undefined between atoms 1611 1610 1612 N2-CM-CM
Angle bend undefined between atoms 1612 1614 1931 CM-N2-H
Angle bend undefined between atoms 1612 1614 1613 CM-N2-CA
Angle bend undefined between atoms 1614 1612 1620 N2-CM-H4
Angle bend undefined between atoms 1614 1613 1621 N2-CA-H5
Angle bend undefined between atoms 1775 1778 1779 CT-CT-CM
Angle bend undefined between atoms 1779 1780 1782 CM-CM-N2
Angle bend undefined between atoms 1779 1781 1784 CM-CM-CM
Angle bend undefined between atoms 1780 1782 1783 CM-N2-CA
Angle bend undefined between atoms 1780 1782 1793 CM-N2-H
Angle bend undefined between atoms 1780 1779 1781 CM-CM-CM
Angle bend undefined between atoms 1781 1783 1785 CM-CA-CA
Angle bend undefined between atoms 1782 1783 1785 N2-CA-CA
Angle bend undefined between atoms 1782 1780 1792 N2-CM-H4
Angle bend undefined between atoms 1784 1786 1787 CM-CA-CA
Angle bend undefined between atoms 1784 1786 1796 CM-CA-HA
Angle bend undefined between atoms 1913 1915 1924 C-CM-HC
Angle bend undefined between atoms 1915 1916 1925 CM-CM-HC
Angle bend undefined between atoms 1916 1917 1918 CM-CA-CA
Angle bend undefined between atoms 1916 1917 1922 CM-CA-CA
Angle bend undefined between atoms 1916 1915 1924 CM-CM-HC
Angle bend undefined between atoms 1917 1916 1925 CA-CM-HC
Angle bend undefined between atoms 1919 1920 1923 CA-C-O
Angle bend undefined between atoms 1921 1920 1923 CA-C-O
MM function not complete
</pre>

Many atoms are not well defined by GaussView, so change them.

Finally you should have change the following atoms

<pre>
< 40 C-CM
< 41 N-N2
< 42 C-CM
< 43 C-CA
< 44 N-N2
< 649 O-O2
< 650 O-OH
< 1610 C-CM
< 1611 N-N2
< 1612 C-CM
< 1613 C-CA
< 1614 N-N2
< 1779 C-CM
< 1780 C-CM
< 1781 C-CM
< 1782 N-N2
< 1783 C-CA
< 1784 C-CM
</pre>

By the following one

<pre>
> 40 C-CC
> 41 N-NA
> 42 C-CV
> 43 C-CR
> 44 N-NB
> 649 O-OH
> 650 O-O
> 1610 C-CC
> 1611 N-NA
> 1612 C-CV
> 1613 C-CR
> 1614 N-NB
> 1779 C-C*
> 1780 C-CW
> 1781 C-CB
> 1782 N-NA
> 1783 C-CN
> 1784 C-CA
</pre>

Now run again the calculation and you have to define manually the other bond lengths and angles missing.

So finally you should have at the end of your input the following MM parameters

<pre>
HrmStr1 S C 227.0 1.810
HrmStr1 NB H 434.0 1.010
HrmStr1 * * 0.0 0.0
HrmBnd1 CV NB H 30.0 120.0
HrmBnd1 CR NB H 30.0 120.0
HrmBnd1 O2 C OH 80.0 125.0
HrmBnd1 CT S C 62.0 100.0
HrmBnd1 S C O 80.0 124.0
HrmBnd1 S C CM 80.0 125.30
HrmBnd1 * * * 0.0 0.0
</pre>

<h5>Input, output fair</h5>

One atom has no atom type : [[Media:wiki_test_3phy_1.gjf]]
[[Media:wiki_test_3phy_1.log]]

One other atom has no atom type : [[Media:wiki_test_3phy_2.gjf]]
[[Media:wiki_test_3phy_2.log]]

Now we know that we have some MM parameters missing : begin to check the atoms types concerned [[Media:wiki_test_3phy_3.gjf]]
[[Media:wiki_test_3phy_3.log]]

First changes operated concerning atoms num 40 41 42 43 44 1610 1611 1612 1613 1614 1780 1782, and then run wiki_test_3phy_4

[[Media:wiki_test_3phy_4.gjf]]
[[Media:wiki_test_3phy_4.log]]

Change atom type of atom 1779 and then run wiki_test_3phy_5

[[Media:wiki_test_3phy_5.gjf]]
[[Media:wiki_test_3phy_5.log]]

Change atom type of atom 1781 and run wiki_test_3phy_6

[[Media:wiki_test_3phy_6.gjf]]
[[Media:wiki_test_3phy_6.log]]

Change atom type of atom 1784 and then run wiki_test_3phy_7

[[Media:wiki_test_3phy_7.gjf]]
[[Media:wiki_test_3phy_7.log]]

So this last should be ok

[[Media:wiki_test_3phy_8.gjf]]
[[Media:wiki_test_3phy_8.log]]

<h5>Atoms charges</h5>
As you can see in order to run an AMBER calculations the charges of all the atoms have to be specified. GaussView is able to specify many charges (the charges correspondent to different amino acid mentioned in the AMBER paper).

<pre>
N-N3-0.159200 0 22.53936399 -2.55554940 12.12649197 L
C-CT-0.022100 0 21.34836399 -2.49054940 11.23349197 L
C-C-0.612300 0 20.19836399 -3.28254940 11.85849197 L
O-O--0.571300 0 20.40836399 -4.18754940 12.64249197 L
C-CT-0.086500 0 21.70036399 -3.08954940 9.87049197 L
C-CT-0.033400 0 21.58536399 -2.01054940 8.79149197 L
S-S--0.277400 0 19.86636399 -1.89254940 8.23649197 L
C-CT--0.034100 0 20.13836399 -0.68754940 6.91449197 L
H-H-0.198400 0 22.76336399 -3.55054940 12.33349197 L
H-H-0.198400 0 23.35136399 -2.10754940 11.65549197 L
H-H-0.198400 0 22.33636399 -2.05454940 13.01449197 L
H-HP-0.111600 0 21.04836399 -1.46054940 11.10849197 L
H-HC-0.012500 0 22.71136399 -3.46954940 9.89449197 L
H-HC-0.012500 0 21.01736399 -3.89554940 9.64449197 L
H-H1-0.029200 0 21.89836399 -1.05954940 9.19849197 L
H-H1-0.029200 0 22.21736399 -2.26954940 7.95549197 L
H-H1-0.059700 0 20.92536399 -1.04054940 6.26349197 L
H-H1-0.059700 0 19.22536399 -0.56354940 6.34849197 L
H-H1-0.059700 0 20.42636399 0.25845060 7.34249197 L
</pre>

But some charges could be missing

<pre>
O-O--0.567900 0 3.95036399 -10.02254940 -0.92050803 L
C-CT--0.182500 0 6.36136399 -7.53054940 -0.90950803 L
H-H-0.271900 0 6.06136399 -7.38354940 -3.47850803 L
H-H1-0.082300 0 6.29536399 -9.54254940 -1.64450803 L
H-HC-0.060300 0 7.17236399 -7.08454940 -1.46550803 L
H-HC-0.060300 0 5.62336399 -6.77654940 -0.68050803 L
H-HC-0.060300 0 6.74436399 -7.95054940 0.00949197 L
N-N- 0 3.48336399 -7.89454940 -1.15250803 L
C-CT- 0 2.08936399 -8.10854940 -0.66650803 L
C-C- 0 1.35636399 -9.01354940 -1.64550803 L
O-O- 0 0.67236399 -9.94054940 -1.26750803 L
C-CT- 0 1.35836399 -6.77154940 -0.58450803 L
C-CT 0 1.44936399 -6.03754940 -1.92550803 L
C-C- 0 0.98036399 -4.59254940 -1.75050803 L
O-O2- 0 1.76836399 -3.70154940 -2.02150803 L
O-OH- 0 -0.30363601 -4.38854940 -1.29150803 L
H-H- 0 3.78536399 -6.99854940 -1.40350803 L
</pre>

That can correspond to some specific amino acid which are not well defined in the AMBER data base, or it can correspond to some unexpected sequences of atoms (in the case of a chromophore for example because it is not a specific amino acid, GaussView is not able to assign some charges).

'''If the charges missing correspond to an amino acid not well defined'''

In this case, have a look at the geometry and at the sequence of atoms, define the amino acid correspond and find in the AMBER paper the charges correspond to this case.

'''If you can not find a similar sequene in the AMBER paper'''
Most of the time it is the chromophore that is not defined (unexpected sequence of atoms).

I do not have an answer for the moment...
It seems that
[http://www.gaussian.com/g_tech/g_ur/k_mm.htm Amber=UnCharged] could be a good keyword, but I did not succeed running the calculation...

We'll see......

<h4>Files</h4>

I put different files here if you want to run some test calculations.

PYP with ONIOM, 2 layers (chromophore + surroundings) : [[Media:oniom_3phy_pyp_ready.gjf]]

<h4>Optimisation</h4>

Now that you have the complete input file, run an optimisation on the structure:

<pre>
#p opt=quadmac oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

Opt=Quadmac is mainly used for problem convergence cases, which does a quadratic step in the coordinates of all the atoms. Such an optimization can use either an updated approximate Hessian for the internal coordinates or an analytically computed Hessian. It computes products of the low-level (MM) Hessian with vectors as needed.

You must specify a ub3lyp method, the default rb3lyp method is not compatible with the system.

This job will terminate without completion:

<pre>
Error termination request processed by link 9999.
Error termination via Lnk1e in /home/gaussian-devel/gaussiandvh01_pgi_725/gdv/l9999.exe at Mon Aug 3 22:16:14 2009.
</pre>

On inspection of the error message and of the optimisation route in GaussView, it is apparent that the job ran out of steps and was fluctuating between two points. To fix this, copy the last geometry into a new input file and specify a smaller step size:

<pre>
#p opt=(quadmac,maxstep=10) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

This job completes normally.

<h5>Input and output fair</h5>
<p>
[[Media:3PHYopt1.gjf]]<br/>
Last geometry: [[Media: ERROR.gjf]]<br/>
[[Media:3PHYopt2.gjf]]<br/>
Final geometry: [[Media:DONE.log]]<br/>

<h4>Frequency</h4>

<h5>Freq</h5>

Check that the optimised structure is really a minimum by running a joint optimisation and frequency calculation:

<pre>
#p opt=(quadmac,maxstep=10) freq oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

The geometry modifies slightly: [[Media:geometry.log]]<br/>

There are no imaginary frequencies, therefore this structure is a minimum (as a tip for ease in .log file navigation, use X11 to open the .log files with the "nedit" command).

<pre>
Low frequencies --- -0.3097 -0.1667 -0.0006 0.0004 0.0006 0.1446
Low frequencies --- 4.5860 5.7688 7.1004
Diagonal vibrational polarizability:
5570.9936676 5790.9109142 7396.2535226
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1 2 3
A A A
Frequencies -- 4.5858 5.7681 7.1004
Red. masses -- 5.6257 7.1713 6.7230
Frc consts -- 0.0001 0.0001 0.0002
%ModelSys -- 0.8136 5.6385 1.5914
%RealSys -- 99.1864 94.3615 98.4086
IR Inten -- 0.1580 0.7010 0.7201
Atom AN X Y Z X Y Z X Y Z
1 7 0.02 0.01 0.03 0.00 0.00 -0.01 0.02 0.01 0.02
2 6 0.02 0.01 0.03 0.00 0.01 -0.02 0.02 0.00 0.04
3 6 0.02 0.01 0.03 0.00 0.01 -0.01 0.02 0.00 0.04
4 8 0.03 0.01 0.04 0.00 0.00 -0.02 0.02 0.01 0.05
5 6 0.02 0.01 0.05 0.00 0.00 -0.02 0.02 0.01 0.05
</pre>

Now copy the same geometry into another input file (to avoid possible corruption with .chk) and run a freq=HPmodes calculation. This keyword gives the vibrational frequency eigenvectors to 5 figures (high precision):

<pre>
Low frequencies --- -0.3101 -0.1651 -0.0004 0.0001 0.0006 0.1435
Low frequencies --- 4.5908 5.7674 7.1007
Diagonal vibrational polarizability:
5570.6337105 5790.9612763 7395.8674830
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1 2 3 4 5
A A A A A
Frequencies --- 4.5906 5.7668 7.1006 7.3648 7.6765
Reduced masses --- 5.6278 7.1696 6.7235 7.0975 7.2004
Force constants --- 0.0001 0.0001 0.0002 0.0002 0.0002
Percent ModelSys --- 0.8203 5.6428 1.5962 0.1342 2.1399
Percent RealSys --- 99.1797 94.3572 98.4038 99.8658 97.8601
IR Intensities --- 0.1579 0.7002 0.7215 0.1148 0.4400
Coord Atom Element:
1 1 7 0.02278 -0.00267 0.02068 0.01019 0.03594
2 1 7 0.01230 0.00471 0.00633 0.05492 -0.00137
3 1 7 0.02519 -0.01023 0.02441 0.07149 -0.00477
</pre>

Note that both of these .log files are very large (in particular the second one which is 915mb with over 12.7 million lines). Consequently your system may crash if too many processes are running at once.

<h5>Freq=ModelModes</h5>

Run a freq=modelmodes calculation on the same geometry.

<pre>
#p freq=(modelmodes) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

This keyword displays modes associated with the small model system in this ONIOM calculation:

<pre>
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
1180 1301 1592
A A A
Frequencies -- 347.2729 400.0784 528.7142
</pre>

Only 31 modes instead of 45?

<h5>Freq=(HPModes,ModelModes)</h5>

Finally run a freq=(HPmodes,modelmodes) calculation:

<pre>
#p freq=(hpmodes,modelmodes) oniom(ub3lyp/6-31g(d):amber=hardfirst)=embedcharge geom=connectivity
</pre>

In comparison, the last two files are much smaller than the first two and therefore easier to work with.

If you compute the calculation with '''Freq=(HPModes,ModelModes) keyword''' you directly get the informations that you need :

* First, you know if you have some imaginary frequencies (no mention of the imaginary frequency if there is no).
<pre>
Full mass-weighted force constant matrix:
Low frequencies --- -727.1752 -603.3920 -441.6366 -348.9122 -336.0877 -330.3771
Low frequencies --- -318.1434 -313.7065 -289.1518
****** 362 imaginary frequencies (negative Signs) ******
Diagonal vibrational polarizability:
52770.5393600 27276.1819608 39738.0499572
NorSel: MapVib= 2181 2332 2360 2364 2521 2847 3397 3653 3874 4190
NorSel: MapVib= 4714 4764 4822 4839 5543 5580 5585 5736 5765 5766
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
2181 2332 2360
A A A
Frequencies -- 781.7748 858.7045 871.9542
Red. masses -- 4.5651 1.7393 1.9629
</pre>

* Second, the precision will be better, because the '''HPModes keyword''' enable you to get 5 figures for the vibrational frequency eigenvectors, which could be useful so as to run some other calculation to et some overlap.

<h5>Input fair</h5>

[[Media:3PHYopt1_1freq.gjf]]<br/>
[[Media:3PHYopt1_1freq2.gjf]]<br/>
[[Media:3PHYopt1_1freq3.gjf]]<br/>
[[Media:3PHYopt1_1freq4.gjf]]<br/>
[[Media:3PHYopt1_1freqMM.gjf]]<br/>

Back to [[ONIOM for biomolecules]]

Back to [https://www.ch.ic.ac.uk/wiki/index.php/ONIOM_tutorial_%28G03%29 ONIOM tutorial (G03)]