[Popgenlunch] abstract for friday popgenlunch (fwd)

Elizabeth Thompson eathomp at u.washington.edu
Thu Apr 16 06:56:40 PDT 2015

---------- Forwarded message ----------
Date: Mon, 13 Apr 2015 17:21:01 -0700 (PDT)
From: Elizabeth Thompson <eathomp at u.washington.edu>
To: Joseph Felsenstein <joe at gs.washington.edu>
Cc: Elizabeth Thompson <eathomp at u.washington.edu>
Subject: abstract for friday popgenlunch


Here it is.

(The results are looking good, but still more to do.)



Gene mapping from inferred identity by descent

The test statistics that inform gene mapping on the basis of marker data
on related individuals are a function of gene identity by descent (IBD). They
depend on the pedigree only insofar as a known pedigree provides prior
probabilities of IBD. In the absence of pedigree information, genetic marker
data provide estimates of IBD that may be used in linkage mappping. However
there are several big challenges. One is the need to infer consistent IBD
jointly among multiple observed individuals. Another is the absence of a
``null distribution'' for the likelihood-based LOD score. These two
challenges were addressed by Chris Glazner in his 2014 PhD thesis, and a paper
based on this work is currently in revision. Referees were positive, but
one requested more background explanations, and other asked for a comparison
with other methods. Hopefully, final results pending and the revision due
next week, I will be able to present both the key ideas of Chris's work and
its application on two (simulated but semi-real) examples, and also the work
I have done in order to meet the requests of the referees.

More information about the PopGenLunch mailing list