[Bioperl-l] PDB ATOM records: name, segid, etc.

[Andrew Dalke]

Kris Boulez:
> [ It's nice to see that people are using these modules and have comments
> on them. I'm aware that PDB writing isn't perfect at the moment. ]

Well, I haven't used the modules.  I've just done way too much structure
work in a previous life :)

> I'll give a lightning talk about Bio::Structure at BOSC 02. We might
> also discus this in more depth there.

How about a BoF on the topic?

Me:
>> It includes a 1.x and 2.x format description, as well as commentary
>> about the XPLOR differences.
>>
>> ftp://ftp.ks.uiuc.edu/pub/group/dalke/UPDB-0.5.tar.gz

Kris:
> Downloaded it. Found the 1.x format description and will have a look at
> it.

As I recall, I typed that in based on my hardcopy from about 10 years
ago.  I can bring a copy of that with me to Canada if you want to see
some history.  :)

Me:
>> The problem with these is that they require good support for:
>>    - bond types/orders
>>    - aromaticity
>>    - chirality
>> 
>> and if you are only used to dealing with PDB files you likely
>> won't know how these should be handled. 

Kris:
> I did a major in chemistry, so these concepts are known to me.

Not meaning any sort of disrespect.  I bring it up because even
in chemistry things like SSSR (Smallest Set of Smallest Rings) isn't
usually discussed.  That is what Daylight uses to compute which
rings are aromatic.  The appropriate reference for this algorithm
is Balducci and Pearlman, from JCICS in 1994.  I have a copy if
you want me to bring that as well.

There is also trickiness for things like "chirality unknown but
opposite to *that* chirality."  And I've heard several different
chemists argue about if a given nitrogen is *really* an aromatic
or not.

That's just for Daylight's model of chemistry.  MDL's is somewhat
different, but I don't know the differences.

BTW, "jcics sssr" is a true Googlewhack.  Even unquoted only one page
comes back in Google search.  But it's only a googlewhack if you,
like me, consider JCICS and SSSR to be common words.

					Andrew
					dalke@dalkescientific.com