[Bioperl-l] automation of translation based on alignment

[Peter]

On Mon, Mar 22, 2010 at 8:51 PM, Chris Larsen <clarsen at vecna.com> wrote:
> Ross, Chris F,
>
> I'd like to just comment on this since we are working in parallel on a
> similar problem. See also the prior thread in archives for Peters work in
> BioPython that I instigated: "Polyproteins, robo slippage, viral
> mat_peptides"

Minor typo - the old thread title was about ribo (ribosomal) slippage:
http://lists.open-bio.org/pipermail/bioperl-l/2009-October/031479.html
http://lists.open-bio.org/pipermail/bioperl-l/2009-October/031484.html
etc

Triggered in part by my discussion with Chris Larsen (off list) about
the biological problem of getting the mature peptide sequences from
GenBank files, Biopython 1.53 ended up with a new method for
extracting the sequence region described by a (complex) location,
e.g. from parsing in an  EMBL/GenBank file. There were several
threads about this, this is perhaps the best summary if anyone is
interested:
http://lists.open-bio.org/pipermail/biopython/2009-November/005813.html
http://lists.open-bio.org/pipermail/biopython/2009-December/005889.html

> This dialog below is just to clarify the science that will guide the
> pseudocode and logic flow would be needed to be built out into a BioPerl
> module. There are plenty of comments on the string mashing required, and its
> a harrowing morass, but heres some other thoughts. Three line item comments
> first, and then some open general ideas for moving this block of concepts
> forward:

Thanks for the update - it sounds like you've got a better understanding
of the complexities now, any some of the reasons why representing
things like mature peptides is tricky (the issue of different cleavage
patterns in different hosts is interesting).

Peter