[Chimera-users] Electrostatic potential molecular surfaces

Wed Sep 11 15:52:47 PDT 2019

Yet another issue I should mention is that in the methods I described, the whole amino acid residue is assigned a given kdHydrophobicity value or polar/nonpolar classification, not distinguishing between which atoms in the residue are more polar than other atoms in the same residue.  To distinguish polar vs. nonpolar at the atom level instead of whole residues, you’d have to use some other program or approach.

I just remembered another web server not related to Chimera. At  least when I tried it many years ago, the results included a column for the apolar area:
<http://curie.utmb.edu/getarea.html>

Elaine

> On Sep 11, 2019, at 3:23 PM, Elaine Meng <meng at cgl.ucsf.edu> wrote:
> 
> Hi Fernando,
> Our resources don’t allow spending much time on consulting with individual research projects, sorry, and I can only answer mostly as it pertains to Chimera as I don’t know the details of parameters developed by other people or other programs.  Also Chimera questions should be sent to this mailing list instead of to me directly so that other people can also benefit, unless private data are discussed or attached.
> 
> You found an old script that is really special-purpose, finding the surface area that is more red than blue after coloring by electrostatic potential.  Note that this would include surface that looks white but has a tiny, tiny tint of pinkishness.  Also, the surface area that is more blue than red is just the complement of the surface that is more red than blue.  In other words, if the surface is 75% red>blue then it is 25% blue>red because very little if any will be pure white.  They always add up to the total surface.  So no, you cannot use these scripts to figure out how much of the surface is polar or nonpolar.
> 
> It  may not matter given the above, but which method, set of charges, and what parameter values to use for electrostatic potential (ESP) calculations are research decisions that I cannot make for you.  There are always tradeoffs, and different people have different opinions, etc.
> 
> I don’t know python, but your generalization to coloring by hydrophobicity by identifying cornflower blue vs.goldenrod might be wrong.  Red vs. blue is simpler since they are RGB color components.  However, even if the color comparison is done correctly, the logical problem is the same, they always add up to the whole surface, and nearly white areas will always be combined with one side or the other. 
> 
> However, you can get the total surface area given some range of hydrophobicity values independent of any coloring.  Hydrophobicity in this case is just a constant lookup value for each type of amino acid, according to this table (kdHydrophobicity column):
> <http://www.rbvi.ucsf.edu/chimera/docs/UsersGuide/midas/hydrophob.html>
> 
> You would have to decide yourself which range of values is polar or nonpolar, but then you could just select all the residues that are less than or more than a certain value, e.g. command:
> 
> select :/kdHydrophobicity>0.0
> 
> Alternatively you could just use command:
> 
> select polar
> select ~polar
> …but you might not agree with what amino acids are considered polar, as shown in a table in this page:
> <http://www.rbvi.ucsf.edu/chimera/docs/ContributedSoftware/resprop/resprop.html#aacat>
> 
> Then if surface is shown you could add up the area for the selected residues, e.g. menu: Tools… Structure Analysis… Attribute Calculator, and in that tool calculate attribute named whatever you want for “molecules” with Formula: 
> 
> sum(areaSAS.residue)
> 
> (… or SES instead of SAS, to be compared with the corresponding total area given in the Reply Log when you show the surface) with option to “Restrict formula domain to current selection” and “Show calculation results in Reply Log" turned on, other options turned off, click Apply.
> <http://www.rbvi.ucsf.edu/chimera/docs/ContributedSoftware/calculator/calculator.html>
> 
> Totally separate and not related to Chimera, but I happen to know of this web server you might be interested in, for comparing ESP of related proteins:
> <https://pipsa.h-its.org/pipsa/>
> 
> Best regards,
> Elaine
> -----
> Elaine C. Meng, Ph.D.
> UCSF Chimera(X) team
> Department of Pharmaceutical Chemistry
> University of California, San Francisco
> 
>> On Sep 11, 2019, at 2:16 PM, Fernando Villa <fer.vdl1928 at gmail.com> wrote:
>> 
>> 
>> Dear Elaine C. Meng, Ph.D
>> 
>> I have some questions about electrostatic potential molecular surfaces and protein comparison.
>> 
>> 1. The following scripts are added, edited from the redarea.py script of link http://plato.cgl.ucsf.edu/trac/chimera/wiki/Scripts, considering that the script allows reading the surface in Å2 colored more red than blue area.
>> 
>> 1. Add scripts:
>> 
>> Command:
>> 
>>> open
>> 
>> -blue_area_03.py
>> 
>> -cornflowerblue_area_03.py
>> 
>> -goldenrod_area_02.py
>> 
>> -red_area_02.py
>> 
>> (the scripts are attached)
>> 
>> 
>> 2. I used the following commands:
>> 
>>> open 4MZ2
>> 
>>> delete solvent
>> 
>>> addh
>> 
>>> addcharge
>> 
>>> Surface
>> 
>> coulombic -10 red 0 white 10 blue
>> 
>> 
>> The electrostatic potential surface is presented by Coulombing coloring Surface:
>> 
>> 
>> 
>> <image.png>
>> I execute the commands:
>> 
>>> redarea #0
>> 
>> red area 3562, total area 5730, ratio 0.6217
>> 
>>> bluearea #0
>> 
>> blue area 2168, total area 5730, ratio 0.3783
>> 
>> 
>> The dominant areas of both colors are obtained.
>> 
>> The following command is input:
>> 
>>> rangecolor kdHydrophobicity min cornflowerblue 0 white max goldenrod
>> 
>> 
>> 
>> <image.png>
>> Next:
>> 
>>> cornflowerbluearea # 0
>> 
>> cornflowerblue area 4109, total area 5730, ratio 0.7171
>> 
>> 
>>> goldenrodarea # 0
>> 
>> Goldenrod area 1621, total area 5730, ratio 0.2829
>> 
>> 
>> So, I believe (I'm not sure):
>> 
>> Blue area + Red area = PSA ß Is correct?
>> 
>>          White area = NPSA ß Is correct?
>> 
>> 
>> So, the real blue and red areas would be (not sure):
>> 
>> 
>> (Redarea * PSA) / Total surf = Real Red area
>> 
>> (Bluearea * PSA) / Total surf = Real Blue area
>> 
>> So:
>> 
>> Bluearea = 1554.6792
>> 
>> Redarea = 2554.3208
>> 
>> 
>> Is this calculation correct?
>> 
>> 
>> Could I apply these calculations to compare them with other proteins of the same family?
>> 
>> For example:
>> 
>> Protein members of the immunoglobulin family.
>> 
>> 
>> 
>> In the Chimera version 1.14, I Download the following programs pdb2pqr-windows-bin64-2.1.1
>> 
>> apbs1.5_win64
>> 
>> Downloaded from the website: http://www.poissonboltzmann.org
>> 
>> Next, I assign charges and radii with the following parameters of PDB2PQR
>> 
>> 
>> <image.png>
>> And I saved the file. Pqr
>> 
>> 
>> <image.png>
>> Later I opened the file 001_B4.pqr with Chimera 1.14
>> 
>> 
>> <image.png>
>> I use the following command:
>> 
>>> surface
>> 
>> And run the APBS interface with the following default parameters:
>> 
>> <image.png>
>> 
>> As a result, the color gradient of the molecular surface appeared as shown:
>> 
>> 
>> <image.png>
>> <image.png>
>> 
>> 
>> However, my question is:
>> 
>> Are these parameters correct using the force field PARSE of PDB2PQR? Or, is it necessary to use a different force field?
>> 
>> 
>> Could I apply the scripts:
>> 
>> -blue_area_03.py
>> 
>> -cornflowerblue_area_03.py
>> 
>> -goldenrod_area_02.py
>> 
>> -red_area_02.py
>> 
>> 
>> to calculate the redarea and bluearea and compare it with other proteins?
>> 
>> 
>> Is there a script that gives me the correct areas of both red and blue colors?
>> 
>> 
>> Sorry for my long email, as I am new to this area. And I really need help.
>> 
>> I apologize for so many questions.
>> 
>> I would greatly appreciate your help.
>> 
>> 
>> Best regards,
>> 
>> Fer.
>> 
>> 
>> 
>> 
>> <blue_area_03.py><red_area_02.py><goldenrod_area_02.py><cornflowerblue_area_03.py>
> 