Algorithms for association between genotype and the resulting physical traits while considering evolutionary history

From the early beginnings of molecular biology physicists have played
a major role in its development starting with Schrödinger's 'What is
Life?' which introduced the idea of an ‘aperiodic crystal’ as a
repository of genetic information. As biology becomes more
quantitative with the advent of technologies such as high throughput
sequencing, physicists have found themselves attracted by the
challenges in biological systems that could be unraveled with the vast
amounts of information now coming online. In this talk, I will
describe one such problem in biology and my approach toward its
solution. A long standing goal in biology is to understand which
genomic regions or genes (genotype) are involved in the expression of
a given observable trait. I will describe two methods that could be
used to correlate a genomic region with a given trait while taking the
evolutionary history of the species or organisms into account. The
first method correlates a binary or discrete character (such as eye
color or presence/absence of a character) with a genotype. The
applicability of this method is demonstrated by its use to find genes
correlated with susceptibility to anthrax in mice.

The limitation of this method to discrete characters is significant as
most characters in nature are continuous (such as height). We develop
a method that can be used to find correlations between a continuous
character and a genotype while taking the evolutionary history into
consideration. Randomization testing is used to assess the significance
of the correlation. I will present a case study using this method
where we find regions in the mice genome that are correlated with
response to a high-fat diet. Comparison of our results with literature