Path: utzoo!utgpu!watserv1!watmath!uunet!zaphod.mps.ohio-state.edu!uwm.edu!bionet!BIOMAC.UNIV-LYON1.FR!cgautier
From: cgautier@BIOMAC.UNIV-LYON1.FR
Newsgroups: bionet.general
Subject: RE codon usage data base
Message-ID: <9102191149.AA06796@genbank.bio.net>
Date: 18 Feb 91 14:56:24 GMT
Sender: daemon@genbank.bio.net
Lines: 18

answer to codon usage database

It is relatively easy to construct codon usage
for a species from data bank as EMBL or GenBank.
Some software do that nearly automaticaly. However
this have a biological significance only for
species having an homegeneous codon usage
(well known case are E. coli (see work of IKEMURA)
Y. saccaromyces (the same author) D. melanogaster
(work of P. Sharp)). For many species like mammals
most viruses no "mean codon usage" exists and
several strategy exists inside the same genome.
This invalidate methods using an overall codon
usage for CDS detection (in my opinion at
least!)

C. Gautier Univ Lyon I
cgautier@biomol.univ-lyon1.fr