[Open Babel] (no subject)

[Xu, H. <Hua...@3d...>]

when I tried to parse the PDB file of a large protein with openbabel, the
program takes an unbearably long time to complete. I followed the thread in
debugger, and realized that the time consumption is concentrated in
assigning implicit valences by matching a set of predefined SMARTS patterns.
Apparently this is prohibitive for large proteins. Shouldn't we take
assignImplicitValence() out of ReadPDB()?

Also, line 54 and 55 in c3d.cpp should be

vector<OBNodeBase*>::iterator i;
vector<OBEdgeBase*>::iterator j;

Otherwise the code cannot be built in Visual Studio .Net.

BTW, as I was looking at the data files (phmodel.txt), I saw some SMARTS
patterns that I couldn't comprehend. What is a ^ in SMARTS? I saw N^3, which
I assume means a nitrogen with 3 explicit connections. But I can't find ^ in
SMARTS manual. Also, what does [#1:1] mean? ':' is used to denote an
aromatic bond in SMILES. But what does it mean in a pattern like O=CO[#1:1]?

Thanks.

Huafeng
Huafeng Xu, PhD                      | Email: hx...@3d...
3-Dimensional Pharmaceuticals, Inc.  | Tel: (609) 655-6978
8 Clarke Drive                       | Fax: (609) 655-6930
Cranbury, NJ 08512