[Biopython] Protein side-chain angles from a PDB file

[Steve Darnell]

Rodrigo,

Perhaps this will get you in the right direction for your
implementation.  It describes the math for calculating dihedrals and
defines the chi angle atom mappings for all of the amino acids.  At
first glance it appears to be correct.  Unfortunately, I don't have any
code to help you out.

http://www.math.fsu.edu/~quine/IntroMathBio_05/torsion_pdb/torsion_pdb.p
df

If you're only interested in calculating the chi angles only once or
twice, you could try MolProbity.

http://molprobity.biochem.duke.edu/ 

Maybe some else knows of another already implemented solution?

~Steve


-----Original Message-----
From: biopython-bounces at lists.open-bio.org
[mailto:biopython-bounces at lists.open-bio.org] On Behalf Of Rodrigo
faccioli
Sent: Monday, September 21, 2009 2:24 PM
Cc: biopython
Subject: Re: [Biopython] Protein side-chain angles from a PDB file

Steve Darnell and Peter Cock,

I apologize your answers. I have complicated my question because I have
not understood very well about these angles. However, I have understood
more.

Thus, I'll retype my original question: How can I calculate the
rotational conformations (rotamers) when I have a PDB file?

Therefore, if I understood better, I need to know which amino acid I'm
working because the number of chi angles vary according to each amino
acids.


Thanks in advance,

--
Rodrigo Antonio Faccioli
Ph.D Student in Electrical Engineering
University of Sao Paulo - USP
Engineering School of Sao Carlos - EESC
Department of Electrical Engineering - SEL Intelligent System in
Structure Bioinformatics http://laips.sel.eesc.usp.br
Phone: 55 (16) 3373-9366 Ext 229
Curriculum Lattes - http://lattes.cnpq.br/1025157978990218


On Mon, Sep 21, 2009 at 12:03 PM, Steve Darnell
<darnells at dnastar.com>wrote:

> Rodrigo,
>
> This is just to expand on Peter's comment.  Your original question 
> implicitly mentioned two types of angles: bond and dihedral angles.  A

> bond angle can be calculated with three atoms, two vectors, and a dot 
> product (the first type mentioned).  When you use the term phi and psi

> angles, you are mentioning dihedral (or torsion) angles (the angle 
> betweeen two planes where the intersection is along the bond of 
> interest).  It's more complicated to calculate, but relatively
straight forward:
>
> http://en.wikipedia.org/wiki/Dihedral_angle
>
> Were you originally asking about how to calculate the torsion angles 
> in the side chain?  These are known as the chi angles and are used for

> defining rotational conformations (rotomers).  I'll stop here since 
> I'm guessing I misunderstood your original question.
>
> Regards,
> Steve
>
> --
> Steve Darnell
> DNASTAR, Inc.
> Madison, WI USA
>
>
> -----Original Message-----
> From: biopython-bounces at lists.open-bio.org [mailto:
> biopython-bounces at lists.open-bio.org] On Behalf Of Peter Cock
> Sent: Monday, September 21, 2009 4:19 AM
> To: Rodrigo faccioli
> Cc: biopython
> Subject: Re: [Biopython] Protein side-chain angles from a PDB file
>
> On Sun, Sep 20, 2009 at 8:24 PM, Rodrigo faccioli < 
> rodrigo_faccioli at uol.com.br> wrote:
> > Hi,
> >
> > I have a doubt about how can I calculate the protein side-chain 
> > angles from a PDB file?
> >
> > I've read the Bio.PDB.Vector evaluates the angles from three atoms.
> > However, my question is how can I  chose these atoms from a PDB
file?
> >
> > I have based on peter cook web-site and I have calculated the 
> > phi-psi angles but I haven't seen about side-chain angles.
>
> http://www.warwick.ac.uk/go/peter_cock/python/ramachandran/calculate/
>
> Its "Cock", not "Cook", but its a common mistake ;)
>
> > Sorry if my question is very basic. However, I'm a computer 
> > scientist novice in chemistry issue.
>
> How are you defining the side chain angle? From memory (without 
> checking the details), you have the protein back bone which includes 
> the "alpha carbon" (CA in PDB files) to which the side chains are 
> attached. I guess you want to measure the angle of the side chain to 
> the C-alpha to (either of
> the) neighbouring backbone atoms.
> The point here is off the top of my head there are at least two 
> possible angles you might be asking about.
>
> But in terms of the code, you'll just need to get the coordinates of 
> the three atoms defining the angle (which probably will be the C-alpha

> and two others), which defines two vectors, then take their dot 
> product and thus get the cosine of the angle. Looking at the Psi-Phi
code should help here.
>
> Peter
>
> _______________________________________________
> Biopython mailing list  -  Biopython at lists.open-bio.org 
> http://lists.open-bio.org/mailman/listinfo/biopython
>
_______________________________________________
Biopython mailing list  -  Biopython at lists.open-bio.org
http://lists.open-bio.org/mailman/listinfo/biopython