Scanning For Selection Using Haplotypes and Multi-Locus Genotypes

A Department of Computational Biology  Seminar

Featuring Dr. Zachary Szpiech, Assistant Professor, Department of Biology, Penn State University

Haplotype-based statistics for identifying genomic regions putatively under positive selection have proven popular for studying adaptation in diverse organisms, although most of these statistics compute summaries of the haplotype frequency spectrum. Here I propose a composite likelihood method based on distortions of the truncated haplotype frequency spectrum and evaluate its performance. In some applications, haplotype phase may not be known or cannot be accurately estimated. In this case, I will show how we can modify this and other popular haplotype-based selection statistics for use with unphased data and demonstrate their effectiveness.