HyPhy message board - Print Page

HyPhy message board
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl Methodology Questions >> How to >> Statistics/info print outs http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl?num=1232717861 Message started by jh26 on Jan 23^rd, 2009 at 5:37am

Title: Statistics/info print outs
Post by jh26 on Jan 23^rd, 2009 at 5:37am

Hi

I have a couple of (perhaps blindingly) obvious questions, but after numerous analyses using Hyphy/Datamoney and looking at the guides and boards, I still haven't found a way to get a print out of:

1) The number of synonymous and nonsynonymous substitutions per codon site, and
2) The overall codon tree length (as opposed to subs/nucleotide).

What are the necessary steps to achieve this? (Dataset of 35 sequences and 210 codon sites).

Also, Hyphy tends to crash after I try and save results to file...is this anything to do with an urgent need to upgrade my (Windows)computer?

Thank you very much in advance,

Jo :-?

Title: Re: Statistics/info print outs
Post by Sergei on Jan 24^th, 2009 at 9:44am

Dear Jo,

Try using Standard Analyses>Miscellaneuos>Phylohandbook.bf>dSdN.bf
and following the prompts.

When the analysis is done, you should get a printout containing this list (obviously with different numbers for your data):

Code (]
Total nucleotide sites :1698
Synonymous sites :373.961
Nonsynonymous sites :1252.14

Total tree lengths:
E[syn):

: 0.0556991
E[non-syn] : 0.0295723
dS : 0.252906
dN : 0.0401025

Sounds like you are looking for E[syn] (syn subs/codon site) and E[non-syn] (non-syn subs/codon site). The total length of the tree is the sum of the two.

S.

Title: Re: Statistics/info print outs
Post by jh26 on Jan 25^th, 2009 at 7:40am

Hi Sergei

That's great, thank you. However is there a way to get a print out of the number of nonsynonymous and synonymous subistitutions for each codon site? I think you can do this using the SLAC method in Datamonkey, e.g:

Codon Observed S Observed NS
Changes Changes
1 1 0
2 3 1
3 1 0
etc

Also, will this depend upon the selection method used (SLAC,REL,FEL), or is there a quick and general way of calculating the above?

Thanks!
Jo

Title: Re: Statistics/info print outs
Post by Sergei on Jan 25^th, 2009 at 5:34pm

Dear Jo,

SLAC will indeed count substitutions for you.
There is no general way to do it, however. One must fit an evolutionary model to the data, infer ancestral sequences and then map mutations to the tree to obtain the two numbers -- they will therefore depend on the model chosen. SLAC is a reasonably good default model - it will perform well unless you have very long tree branches (>1 sub long), when SLAC can undercount the number of substitutions per site.

Best,
Sergei