> I am trying to work out the replacement/substitution ratios in the
> CDR and FR regions of some human autoantibody sequences. The problem
> is how to score a situation where two or three bases in a codon for a
> given amino acid are changed. Is it normal practice to consider each
> mutation separately for the effect it would have had if the other
> bases were in germline configuration? Or can you try to score on a
> single codon basis? If you are using these figures as an index of
> antigen-driven selection then the approach used is important as the
> values will be different and the former contains assumptions about
> the order of mutations which is unknown.
> Any comments or have I missed something?
> Mark Hexham
> Sheffield UK
>I would suggest you look up: Dikes and King, 1979 Nature 281:605 and
also Nature 1987, 328:805-811. There is some discussion about R/S
ratios in these papers. If you are interested in frameworks and CDRs
you should also look up a paper by Kirkham et. al. in EMBO late last
year or this year.
I hope this helps. It should at least get you started.
jim george
txpljfg at CVSR.UAB.EDU
