BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:MRC Biostatistics Unit Seminars
SUMMARY:&quot\;A Bayesian analysis of microbiome data&quot
 \; - Dr Sergio Bacallado\, University of Cambridge
DTSTART;TZID=Europe/London:20161129T143000
DTEND;TZID=Europe/London:20161129T153000
UID:TALK67164AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/67164
DESCRIPTION:Techniques based on sequencing 16S ribosomal DNA h
 ave been used for several years to characterise mi
 crobial communities. We introduce a Bayesian nonpa
 rametric analysis of dependent discrete distributi
 ons which can be applied to the analysis of these 
 experiments. The procedure deals effectively with 
 rare species without the need for truncation or ra
 refaction\, and the model makes it possible to inf
 er species co-occurrence patterns that are usually
  observed in these datasets\, unlike other Bayesia
 n approaches such as the Dirichlet-Multinomial mod
 el without prior dependence. The dependence betwee
 n distributions is expressed by latent features\, 
 which makes the model especially suited to the joi
 nt analysis of multi-omic experiments which might 
 generate\, for example\, metabolic profiles\, RNA 
 expression data\, or proteomics data in addition t
 o the species-level characterisation of a 16S sequ
 encing experiment. We describe a simple approach t
 o visualise the posterior uncertainty in ecologica
 l ordinations typically applied in microbiome stud
 ies.
LOCATION:Large  Seminar Room\, 1st Floor\, Institute of Pub
 lic Health\, University Forvie Site\, Robinson Way
 \, Cambridge
CONTACT:Alison Quenault
END:VEVENT
END:VCALENDAR