On Mon, 26 Aug 2002, Richard Gillilan wrote:
> Maybe soon I can get some more time to look into this issue again.
> I've been totally preoccupied with other stuff.
> I really need to get it working for proteins in order to get a paper out.
> I had made a lot of hacks to an older CVS version of the code so that it
> almost got everything right for all the amino acid residues. Still there
> a couple of problems I didn't solve. Geoff was going
> to make a lot of changes to that part of the code last we spoke and I
> don't want
> to check in any of the hacks since these are only temporary fixes. If
> you like
> I can tell you what I did.
That would be cool actually ; I'm thinking vaguely of forcing a
template typing all ALA's are of a certain bond order HIS comes in one of
3 to be checked for etc etc ; maybe patty everything and go back and
smart-patch against a residue set afterwards.
Geoff, here is the patty bug (in patty.cpp, not in patty.h)
it is in the last bit of the read_rules() routine : At least this is
what I need to get the Primer.html example code etc to run properly.
Perhaps there is an initialization of this array somewhere that I didn't
if (buffer != '#')
if (vs.size() >= 2)
_sp[i] = new OBSmartsPattern; // DEBUG AVI
printf("running push back\n");
printf("Done with push_back\n");
>At the moment, the bond order typer is pretty stupid. As Richard
>mentioned, I'm in the process of improving it. The first step, which I
>may get to this afternoon is to add a full set of functional group typing
>and fix a few minor bugs. But the real problem is the aromatic group
>perception pass, which will take me some time unless I can figure out how
>to hack into the current typers.
Okay cool; let me play with this a bit more and see what weird
things i keep hitting. Its odd but I dont seem to catching ring errors so much
that occasionally it seems like the odd hydrogen seems to be bonded to the
wrong atom :(