[Chimera-users] Computing 'basic' structure characteristics

Elaine Meng meng at cgl.ucsf.edu
Sun Oct 9 09:53:16 PDT 2016


Hi Joe,
Chimera has “addcharge” or “pqr” commands for associating atoms with charges and “apbs” to run Poisson-Boltzmann electrostatic calculations and return an electrostatic potential (ESP) map for surface coloring.  The latter two actually call welb services. There is also a “coulombic” command for coloring surfaces by Coulombic (instead of Poisson-Boltzmann) ESP.

These are also all available as tools with graphical interfaces, but you mentioned scripting so I listed the commands.

However, I don’t know what you mean by surface charge, but unless you meant ESP Chimera does not calculate it, nor does it predict pKa.  As Jaime said, you can use ProPka (he also gave an URL… hey Jaime, we’re excited to hear you might write a Chimera interface for this!).

If you just want to look at predominance of secondary structures, you would not use those integers which mean first strand, second strand, first helix, etc.  Instead you could just see how many residues are already assigned as strand, helix, and coil.  There aren’t commands to do this directly, only some very cumbersome approaches that I imagine are vastly inferior to using Python (like “select strand” and then writing a list of all those residues with “writesel” etc.).

I can’t help with the Python side, this is just the perspective from the commands side. All the commands I mentioned are documented,
<http://www.rbvi.ucsf.edu/chimera/docs/UsersGuide/framecommand.html>

Best,
Elaine
-----
Elaine C. Meng, Ph.D.                       
UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
Department of Pharmaceutical Chemistry
University of California, San Francisco

On Oct 8, 2016, at 8:38 AM, Healey, Joe <J.R.J.Healey at warwick.ac.uk> wrote:

> Hi again Chimera Team,
> 
> I have 2 questions, one specific and one quite general. If you see the emails below, I was speaking with Jaime about computing basic structure characteristics (pKa, net surface charge, proportions/predominance of 2ary structure etc).
> 
> The general question I have is what suggestions might you have for functions within chimera capable of computing any or all of these features? (ideally commandline based so that I can run this over many PDB structures).
> 
> The specific question that pertains to the below is that Jaime mentioned each residue has an int value that corresponds to a particular structure designation, ints from -1 to 6. Could you let me know what the correspondence is between these values and actual structure?
> 
> Joe Healey
> 
>                                         
> M.Sc. B.Sc. (Hons)
> PhD Student
> MOAC CDT, Senate House
> University of Warwick
> Coventry
> CV47AL
> Mob: +44 (0) 7536 042620  |  Email: J.R.J.Healey at warwick.ac.uk
> 
> Jointly working in:
> Waterfield Lab (WMS Microbiology and Infection Unit)
> and the Gibson Lab (Warwick Chemistry)
> 
> Twitter: @JRJHealey  |  Website: MOAC Page
> 
> 
> From: Jaime Rodríguez-Guerra <jaime.rodriguezguerra at uab.cat>
> Sent: 08 October 2016 09:55
> To: Healey, Joe
> Subject: Re: [Chimera-users] Recursive structure matching and acquisition of descriptive numbers
>  
> Hi again Joe!
> 
> No problem at all; I'll be pleased to help you.
> 
> For charges, I think Chimera can calculate them as part of the Surface/Binding analysis> Dock Prep wizard. You also have some information about 2nd structure inside chimera.Residue objects, specifically ssId attribute. However, these are int values, and I don't really know which ss is each (ranging from -1 to 6, as I have observed).
> 
> For pKa, you can use PropKa (https://github.com/jensengroup/propka-3.1), and then I guess you'll be able to get the pI somewhere around the code. I've been designing a GUI to use this in Chimera, but it's not yet ready... Sorry about that!
> 
> I don't know if you need anything else... but I think the Python community will have you covered one way or another! BioPython.PDB might be useful too (http://biopython.org/wiki/The_Biopython_Structural_Bioinformatics_FAQ).
> 
> Cheers,
> Jaime.
> 
> 
> 2016-10-06 15:57 GMT+02:00 Healey, Joe <J.R.J.Healey at warwick.ac.uk>:
> Hi again Jaime,
> 
> Hope you don't mind but I had a couple of questions for you. You may have figured out from all the previous correspondance that I have a load of protein structure simulations. The RMSD script you helped me with was to sort of 'broadly' gauge whether the simulations looked realistic or not.
> 
> For some of them, I'd like to build up a chart of their tertiary structure properties based on the simulations. Things like charge/hydrophobicity, isoelectric point, 2ndary structure dominance and so on.
> 
> Could I ask you for any advice you might about what tools are available to analyse PDBs? I'm sure chimera is probably capable of it too. I've mostly found webservers at the moment for specific tasks, but I'd like to do most of the analysis locally. My group is full of bioinformaticians, but I'm the only one who really looks at proteins so I need the advice of a computational chemist I think!
> 
> Cheers,
> 
> Joe
> 
> 
> Joe Healey
> 
>                                        
> M.Sc. B.Sc. (Hons)
> PhD Student
> MOAC CDT, Senate House
> University of Warwick
> Coventry
> CV47AL
> Mob: +44 (0) 7536 042620  |  Email: J.R.J.Healey at warwick.ac.uk
> 
> Jointly working in:
> Waterfield Lab (WMS Microbiology and Infection Unit)
> and the Gibson Lab (Warwick Chemistry)
> 
> Twitter: @JRJHealey  |  Website: MOAC Page
> From: Healey, Joe
> Sent: 23 September 2016 11:51:40
> To: Jaime Rodríguez-Guerra
> Cc: Eric Pettersen; chimera-users at cgl.ucsf.edu
> 
> Subject: Re: [Chimera-users] Recursive structure matching and acquisition of descriptive numbers
>  
> Not at all! Much appreciated, as I think I mentioned, this is my first go at writing anything remotely complex in python (and not a bad turn out for about a week I don't think...)
> 
> A lot of the mess comes from copying functional bits from StackOverflow in the interests of just getting it working first, and making it pretty second as I'm sure you can appreciate (you'd hate to see my LaTeX preamble)!
> 
> Not needed to close the file with 'with' is useful to know (takes some of the headache out of knowing where to place the close).
> 
> I'll certainly make the tidy ups you've suggested, for my own learning.
> 
> 
> Thanks again,
> 
> Joe :)
> 
> Joe Healey
> 
>                                        
> M.Sc. B.Sc. (Hons)
> PhD Student
> MOAC CDT, Senate House
> University of Warwick
> Coventry
> CV47AL
> Mob: +44 (0) 7536 042620  |  Email: J.R.J.Healey at warwick.ac.uk
> 
> Jointly working in:
> Waterfield Lab (WMS Microbiology and Infection Unit)
> and the Gibson Lab (Warwick Chemistry)
> 
> Twitter: @JRJHealey  |  Website: MOAC Page
> From: Jaime Rodríguez-Guerra <jaime.rodriguezguerra at uab.cat>
> Sent: 23 September 2016 10:56:38
> To: Healey, Joe
> Cc: Eric Pettersen; Jaime Rodríguez-Guerra; chimera-users at cgl.ucsf.edu
> Subject: Re: [Chimera-users] Recursive structure matching and acquisition of descriptive numbers
>  
> Hi Joe!
> 
> You're welcome! Glad to see you got it working. I've reviewed your
> code and would want to point out some details that might be helpful to
> you in the future, though. These are mostly style-related things, but
> those will contribute to a cleaner reading experience!
> 
> 1. You don't need backslashes if you're inside parenthesis. So, in all
> those parse_argument lines, you can safely delete them. Related to
> this, you can wrap imports with parenthesis, so that backslashes are
> not needed.
> 2. Tuple unpacking does not need parenthesis in you're dealing with
> 1D-tuples. Only needed if you are dealing with more than one dimension
> (ie, (animal, fruit), sport = [['tiger', 'pear'], 'soccer']
> 3. Since you are opening files with the "with" context manager, you
> don't need to manually close the file later. It's closed automatically
> as soon as you leave that with block.
> 4. Lines 166-170 could be replaced with a (cleaner, but not
> necessarily more performant) glob.glob() call. Check it out to see if
> it satisfies your requirements.
> 5. While you are at it, take a look at the PEP8 docs and the Google
> Python style guide. You don't need to follow all the rules, but they
> are really helpful in bringing consistency to your own style!
> 
> That's it! Hope you don't mind these pieces of advice :)
> 
> Cheers,
> Jaime.
> 
> 2016-09-23 11:27 GMT+02:00 Healey, Joe <J.R.J.Healey at warwick.ac.uk>:
> > That was exactly what I was after thanks! I figured the object orientation
> > should provide that somewhere but was barking up slightly the wrong tree!
> >
> >
> > Thank you all for all your help - definitely wouldn't have been able to do
> > it without you!
> >
> >
> > If it's of any interest, or it helps for any future questions where you
> > might want to refer back to code segments, I've put the
> > more-or-less-finished script in this paste:
> >
> >
> > http://pastebin.com/xJNSWJGq
> >
> >
> > Joe Healey
> >
> >
> > M.Sc. B.Sc. (Hons)
> > PhD Student
> > MOAC CDT, Senate House
> > University of Warwick
> > Coventry
> > CV47AL
> > Mob: +44 (0) 7536 042620  |  Email: J.R.J.Healey at warwick.ac.uk
> >
> > Jointly working in:
> > Waterfield Lab (WMS Microbiology and Infection Unit)
> > and the Gibson Lab (Warwick Chemistry)
> >
> > Twitter: @JRJHealey  |  Website: MOAC Page
> > ________________________________
> > From: Eric Pettersen <pett at cgl.ucsf.edu>
> > Sent: 22 September 2016 20:54:35
> > To: Jaime Rodríguez-Guerra
> > Cc: chimera-users at cgl.ucsf.edu; Healey, Joe
> > Subject: Re: [Chimera-users] Recursive structure matching and acquisition of
> > descriptive numbers
> >
> > Thanks Jaime — exactly right.  I realized that the “atoms1, atoms2” in my
> > original example didn’t really give any indication of which atoms were
> > which, which is why in my later examples I changed to "simAtoms, refAtoms”,
> > so it’s actually “atoms2” that are the reference atoms.
> >
> > —Eric
> >
> >> On Sep 22, 2016, at 11:45 AM, Jaime Rodríguez-Guerra
> >> <jaime.rodriguezguerra at uab.cat> wrote:
> >>
> >> Hi!
> >>
> >> Since you are getting lists of atoms back, you only need to do some
> >> attribute access (rather than method calling)!
> >>
> >> Every chimera.Atom object has an attribute called molecule: a
> >> reference to its parent chimera.Molecule. chimera.Molecule objects
> >> have an attribute called 'name', but sometimes this is not very
> >> informative (depends on your input), so you need to resort to the pdb
> >> filename, stored in the first element of the openedAs tuple.
> >>
> >> In code, this looks like this:
> >>
> >> for atoms1, atoms2, rmsd, fullRmsd in match(CP_BEST, [ref, sims],
> >> defaults[MATRIX], "nw", defaults[GAP_OPEN], defaults[GAP_EXTEND]):
> >>    molecule1 = atoms1[0].molecule # any atom from the list will do
> >>    molecule2 = atoms2[0].molecule
> >>    print molecule1.name, "\t", molecule2.name, "\t",  rmsd #
> >> molecule1.openedAs[0] will also work here
> >>
> >> I don't know if atoms1 comes consistently from the reference molecule,
> >> but I'd guess it does.
> >>
> >> Hope it helps!
> >> _______________________________________________
> >> Chimera-users mailing list: Chimera-users at cgl.ucsf.edu
> >> Manage subscription:
> >> http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
> >
> 
> _______________________________________________
> Chimera-users mailing list: Chimera-users at cgl.ucsf.edu
> Manage subscription: http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users




More information about the Chimera-users mailing list