BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Application of Bayesian model averaging and population Monte Carlo
  to inference from metagenomic mixture - Morfopoulou\, S (University Colle
 ge London)
DTSTART:20140328T143000Z
DTEND:20140328T150000Z
UID:TALK51680@talks.cam.ac.uk
CONTACT:Mustapha Amrani
DESCRIPTION:Co-author: Vincent Plagnol (University College London Genetics
  Institute) \n\nFor many practical applications\, for example to uncover t
 he pathogen that caused an infection after the acute phase\, very deep sho
 rt read sequencing can be effective provided that we can reliably assign s
 hort sequencing reads to species. This problem of assignment of reads to s
 pecies is complicated by the fact that\, in the absence of very large cont
 igs\, most short reads reads match to multiple species. This is essentiall
 y a mixture model\, where the complete knowledge of all species present in
  the mixture provides information about the assignment of each read indivi
 dually. However\, metagenomic data analysis rarely formulates the problem 
 in these terms because the very large number of potential species typicall
 y makes the inference intractable. Here\, we propose a Bayesian model aver
 aging strategy designed to explore the high dimensional space of species p
 resent in a metagenomic mixture. We use approximate Bayesian computation a
 nd a Monte Carlo strategy to implement the search o f the most appropriate
  mixture models. Owing to the computationally intensive aspects of the wor
 k\, we used a population Monte Carlo Markov Chain to leverage the use of p
 arallel computing. We find that the methodolgy is effective to provide a f
 ull Bayesian inference for samples with > 10M reads\, hence providing inte
 rpretable Bayes Factors and posterior probabilities for practical problems
  that regularly arise in a clinical context.\n
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR