The email below makes some good points. I can answer that "ligand" etc. are surface categories and not atom specs, but it does beg the question of why we couldn't make them usable as atom specs. The possible answer to the second item is that so far, we have focused more on handling correct PDB files, not various pathological cases. However, I am not sure what is supposed to happen in a correct PDB file when part of a chain is lacking density (e.g. a chain with residues 1-20 lacking density for residues in the middle...is there supposed to be a TER card in that case?). Comments on these issues welcome. Thanks, Elaine --------------------------------------------------------------------
From boas@stanford.edu Thu May 2 14:10:18 2002 Date: Thu, 02 May 2002 14:10:13 -0700 To: Elaine Meng <meng@cgl.ucsf.EDU> From: "F. Edward Boas" <boas@stanford.edu> Subject: Re: Chimera issues/questions
Hi Elaine, I wasn't expecting such a quick reply ... thanks! I have a couple more suggestions: * Why can you "surface ligand", but not "select ligand"? * Automatic chain break detection would be really helpful. Currently, Chimera connects all the residues in a chain, even when it's clear from the bond length that there must be a chain break. Thank you! Cheers, Ed ________________________________________________________________________ F. Edward Boas Phone: 650-787-1688 P.O. Box 17890 Email: boas@post.harvard.edu Stanford, CA 94309 Web: http://www.stanford.edu/~boas
On Thursday, May 2, 2002, at 02:42 PM, Elaine Meng wrote:
The email below makes some good points. I can answer that "ligand" etc. are surface categories and not atom specs, but it does beg the question of why we couldn't make them usable as atom specs.
You can of course use the far more awkward "@/surfaceCategory=ligand" atom spec. I don't think I have any objection to the emulator trying to interpret otherwise-unrecognized atom specs as surface categories, but we should probably discuss the ramifications and alternatives before making such a change.
The possible answer to the second item is that so far, we have focused more on handling correct PDB files, not various pathological cases. However, I am not sure what is supposed to happen in a correct PDB file when part of a chain is lacking density (e.g. a chain with residues 1-20 lacking density for residues in the middle...is there supposed to be a TER card in that case?). Comments on these issues welcome. Thanks,
There won't be a TER card because there is no chain break. The "long bond" is actually informative in that it indicates missing data, rather than a true break. I think the current behavior should be retained, since it is trivial to remove the long bonds with the "longbond" command. --Eric
participants (2)
-
Elaine Meng -
Eric Pettersen