Approximate Bayesian Inference of Bacterial Population Trees with Next-Generation Sequencing Data

• Activity Number: 121
• Type: Topic Contributed
• Date/Time: Monday, August 2, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Alexander V. Alekseyenko*+ and Marc Suchard
• Companies: New York University and University of California, Los Angeles
• Keywords: phylogenetic inference ; Markov chain Monte Carlo ; next-generation sequencing

Abstract

Next-generation sequencing techniques, such as Roche 454 Pyrosequencing, are capable of probing thousands of individual bacterial sequences from a single sample for example by using universal 16S rRNA probes. Processing many such samples through a phylogenetic pipeline in order to arrive at a classification of clinical phenotypes based on bacterial population structure can be a slow process, especially if one resorts to highly computationally intensive Bayesian inference techniques involving Markov chain Monte Carlo sampling. Luckily, many sequences in such samples can be treated as coming from the same operationally defined 'species' (ODS), such that we can be only concerned with the relationships of ODS to each other and not individual sequences. We develop efficient approximate Bayesian methodology for inference of between ODS phylogenies by averaging the structures within ODS.