Inferring Selection from Large-Scale Next-Generation Sequencing Data

• Activity Number: 2
• Type: Invited
• Date/Time: Sunday, August 1, 2010 : 2:00 PM to 3:50 PM
• Sponsor: WNAR
• Author(s): Rasmus Nielsen*+
• Companies: University of California, Berkeley

Abstract

New next-generation sequencing technologies have facilitated cheap and fast genome-wide sequencing allowing analysis at population genetic data at a genome-wide scale. However, valid analyses of such data requires that the special properties of the data are taken into account. We can model the data as arising from a process in which a Poisson distributed number of sequences is obtained from each individual in each gene, and in which the mean of the Poisson varies both among genes and among individuals. In addition, sequencing errors occur at a relatively high rate and is determined by the local sequence. We show how traditional population genetic methods for analyses can be modified to accommodate this process, and show applications to human genome-wide data, and data sets from other organisms.