[Owl-devel] HETATMs support
stehr at molgen.mpg.de
Thu May 19 11:52:42 EDT 2011
Awesome! Thanks a lot Jose! Take that Biojava! :)
On Thu, May 19, 2011 at 4:44 PM, Jose M. Duarte <jose.m.duarte at gmail.com> wrote:
> So more major changes to owl. This time we get a long awaited new feature:
> support for HETATOMs and nucleotides.
> Basically the whole content of PDB entries will be read: standard amino
> acids of protein chains (as before), non-standard amino acids within protein
> chains (HETATMs in protein chains), nucleotide chains and HETATMs that are
> not part of any chains (ligands, cofactors, buffer molecules or any other
> molecule). I've excluded the reading of waters for now as for most uses
> waters are not needed so we can save some memory there. If needed we can
> change that easily, it's just a flag that ignores waters.
> The main change has been in the Residue class. Now we have a Residue
> interface and 3 implementations of it: AaResidue for standard amino acids
> (what used to be called simply Residue), NucResidue for nucleotide residues
> and HetResidue for all other molecules.
> In the process of doing this I've learned that the PDB as usual decided to
> do strange things with the format. Basically there are two kind of chains in
> PDB entries: polymer (poly-peptides or nucleic acid chains) and non-polymer
> (ligands and so on). Classically the non-polymer chains were not treated
> separately but rather they were assigned chain codes that put them together
> with the polymer chains to which they were related: ligands with the chain
> they bind to and so on. Now in the CIF format they decided to then assign
> separate chain codes (CIF chain codes) to the non-polymeric chains, which in
> my opinion makes a lot of sense. But for some reason they decided to keep
> the old chain codes in PDB files.
> What I've tried to do with PDB files is to treat the non-polymer chains
> separately assigning them chain codes trying to keep this assignment as
> close as possible to the CIF assignment. To distinguish both kind of chains
> (polymeric/non-polymeric) use the new method PdbChain.isNonPolyChain().
> I've been trying to track all calls to methods that could be affected and
> have tried to keep compatibility. But of course this is a major change that
> can introduce some bugs, so we will need some time before things are more
> stable again. I've created a tag to mark the code before this change
> Anyway hopefully the change is for the better as now we will be able to do a
> lot more with the PDB data.
> Owl-devel mailing list
> Owl-devel at bioinformatics.org
More information about the Owl-devel