UNC Computational Genetics

I was thinking about the problem of SNPs for which we have an equal number of 0s and 1s. Is it necessary to always select BL6 as the reference? The reason that I don’t want to do this is because the user may one day want to analyze a set of strains without BL6 in it. It seems that it should be equivalent to simply select the first strain in the data set and set its allele = 0 and then propagate that change to the other strains. This will achieve the goal of removing the repeated SDPs. Is there some subtlety that I’m missing here or should this work?

FYI, in the 32 McClurg paper strains, I had 64790 unique SDPs before ths correction and now have 64157 unique SDPs with the correction. In the 8.3M SNP data set, I now get 32768 unique SDPs.

Email Leonard any agenda items

Email Leonard any agenda items.

Email Leonard any agenda items.

Email Leonard any agenda items.

This week Lynda will discuss her tool for selecting optimal validating F2 crosses when given a set of QTLs identified in a panel of RIL-strain test trials.

Prof Wei Wang and I will discuss new approaches to linkage and association mapping of QTL. Wei will discuss approaches for correlating phenotypes to phylogenetic trees, and assessing the quality of their match. I will discuss approaches that resample at target positions (e.g. genes) rather than being limited to evaluations only at marker positions, by taking advantage of genome interval assignments.

A related paper can be found at:

Whole genome association mapping by incompatibilities and local perfect phylogenies

Email Leonard any agenda items

1) Lynda Yang will discuss fast approaches to permutation testing

1. Introductions and beginning of semester stuff
2. Dan Gatti will discuss his and Andrey’s FastHap tool