[gmx-developers] Re: gmx-developers Digest, Vol 14, Issue 9

[Michael Shirts]

Hi, all-

Still trying to figure out the logic of this code that I posted before
-- I feed in a pdb with hydrogrens already placed (I'm trying to
compare to another code, and I need both programs to take the same
input), and it runs through this code to assign which topolgy to use. 
This seems very odd, because it checks to see if the N's to be
protonated are involved with H-bond acceptors or not.  This has
nothing to do with assigning the topologies.

Can somebody explain why this code was used to assign the correct
histidine topology?  This seems to be an algorithmic bug.  I can try
to fix it, but I'd prefer not to until I understand why this is here
in the first place, as I might be misunderstanding what it's doing.

Thanks,
Michael

> I'm working with the hizzie.c code, and it seems to me that it's not
> working correctly in the assignment of a given HIS residue to HIS A or
> HIS B.  Have other people had this problem?  I'm discovering that in
> the code:
> 
>           bHDd = chk_hbonds(nd1,pdba,x,acceptor,hbond,xh1,angle,dist);
>           chk_hbonds(nd1,pdba,x,donor,hbond,xh1,angle,dist);
>           bHEd = chk_hbonds(ne2,pdba,x,acceptor,hbond,xh2,angle,dist);
>           chk_hbonds(ne2,pdba,x,donor,hbond,xh2,angle,dist);
> 
>           if (bHDd) {
>             if (bHEd)
>               type = ehisH;
>             else
>               type = ehisA;       }
>           else
>             type = ehisB;
>           fprintf(stderr,"Will use %s for residue %d\n",hh[type],hisnr+1);
> 
> If I put through a PDB file with a HD1 on ND1 (i.e. it should assign
> ehisA), I get bHDd=bHEd=0, which defaults to ehisB.
> 
> First, this should give an error, (it's assuming that if bHDd =0, then
> bHDe=1), which is not the case.  That, I can fix.  Secondly, it should
> set  bHDd=1 if HD1 is bonded to ND1.  After looking through
> chk_hbonds, I can't quite figure out how to fix this :).
> 
> Thanks,
> Michael
> 
> 
> ------------------------------
> 
> _______________________________________________
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-developers
> 
> 
> End of gmx-developers Digest, Vol 14, Issue 9
> *********************************************
>