Title: "Use of Allele Specific Expression (ASE) to fine map putative QTL associated with disease"
Speaker: W.M. Muir, Department of Animal Sciences, Purdue University

Place: Rawls (RAWL) Hall 1086
Date: October 1, 2013; Tuesday
Time: 4:30pm

Abstract:
Locating causative QTL associated with traits is the holy grail of mapping methods but remains elusive for a variety of reasons, including for one or more of: lack of power, large LD blocks, and unknown population structure. A new tool emerging to fine map QTL associated with treatments or conditions, such as disease, is allele specific expression (ASE). ASE is able to overcome issues with LD because they are cis-acting factors that act on regulatory elements. We used ASE to find QTL associated with Marek's disease in poultry, a T cell lymphoma caused by the highly oncogenic alphaherpesivurs, MDV. RNAseq was performed in replicated F1 birds produced by crossing inbred lines, and either exposed to MDV or not. We then tested for intra-locus differential expression (DE) that changed with exposure (E), i.e. DE x E interaction. The resulting loci are putative QTL associated with MD, we identified 4,500 SNPs in 3,700 genes. To validate the QTL we designed a SNP chip with 9,000 SNPs, 1,800 were ASE SNPs showing greatest interaction, another 3,996 SNPs identified by signatures of selection (SS), and 3,097 were SNP chosen at random. These were used to assay ASE in 1,000 birds exposed to MD from a resource population consisting of the F7 generation of the above cross. The phenotype was presence/absence of MD at 56 days as determined by necropsy. The resulting data was subjected to genome wide association analysis (GWAS) using the difference subsets of SNPS, and also classic pedigree/REML to determine how much of the genetic variation in the trait could be attributed to the sets or pedigree. Results showed that ASE SNPs could account for 83% of the genetic variation in the trait while Random could account for 65%. On a per SNP basis, the ASE SNPs accounted for 2.5 x more variation, suggesting they were tagging causative SNPs while the Random SNPs were predicting relationships, i.e. the pedigree. GWAS verified that the SNPs with greatest effect were identified by ASE. As a final step, we are selecting birds based on their ASE genotypes to establish a cause-effect relationship. All previous studies have inferred putative causation from association of genotype with phenotype, we hope to induce a phenotypic change through specific genotype changes.

Associated reading:
1. MacEachern S., WM Muir, SD Crosby, and HH Cheng 2012. Genome-wideidentificationandquantificationofcis-andtrans-regulatedgenesrespondingtoMarek'sdiseasevirusinfectionviaanalysisofallele-specificexpression. Frontiers in Genetics. 2:1-11.

2. PJ Wittkopp, BK Haerum and AG Clark. Models of regulatory divergence. Evolutionary changes in cis and trans gene regulation Nature 430, 85-88(1 July 2004) doi:10.1038/nature02698



Click here www.stat.purdue.edu/~doerge/BIOINFORM.D/FALL13/sem.html for a full schedule of BIOINFORMATICS SEMINARS, past and present.