[Bioperl-l] RE: population genetics

[Freimuth,Robert]

> >>Hi all,
> >>
> >>I would like to go for an LD viewer and infer haplotypes based on EM
> >>model. I was wondering is some on the list is doing 
> something like that.
> >>
> >>Cheers
> >>Pedro.

> >     I don't know about LD (linkage disequilibrium?) and EM models on
> >bioperl, but there are probably implementations for various 
> applications
> >of it on R statistics software (I know there is for QTL analysis). I
> >recently came across this module to control the R 
> interpreter from Perl:
> >
> >      
> http://search.cpan.org/~gmpassos/Statistics-R-0.02/lib/Statistics/R.pm
> >   
> >     Seeing that R is also Open Source, it might be 
> interesting to use it
> >with BioPerl.
> >
> >     -Paulo Almeida
> >

Regarding LD, I recently came across a program (written in perl) called
ldSelect that calculates r-squared and selects tag SNPs
(http://droog.gs.washington.edu/ldSelect.html).  I haven't played with it
yet, but it might do what you need.  I do not know of any existing perl
programs that calculate D', although there are some available in other
languages.

There are also a few programs available to infer haplotypes.  A popular one,
which uses Bayesian methods, is called PHASE
(http://www.stat.washington.edu/stephens/software.html).  It's free for
non-commercial use but I don't think it is open source.  I use a different
program (also called PHASE) to infer haplotypes using the EM algorithm and
calculate pairwise D' values.  It was originally written by Dan Schaid in S+
and has since been convered to R so it can run on our web server
(http://ilya.wustl.edu/~pgrn/programs.html).

HTH,

Bob