BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Coalescent-based Species Tree Inference from Gene Tree Topologies 
 Under Incomplete Lineage Sorting by Maximum Likelihood - Wu\, Y (Connectic
 ut)
DTSTART:20110623T092000Z
DTEND:20110623T094000Z
UID:TALK31845@talks.cam.ac.uk
CONTACT:Mustapha Amrani
DESCRIPTION:Incomplete lineage sorting can cause incongruence between phyl
 ogenetic history of genes (the gene tree) and that of the species (the spe
 cies tree)\, which can complicate the inference of phylogenies. Developing
  robust computational inference approaches is currently of interests in st
 udying incomplete lineage sorting. In this talk\, I will present a new coa
 lescent-based algorithm for inferring species tree with maximum likelihood
 . I will first describe an improved method for computing the probability o
 f a gene tree topology for a species tree\, which is much faster than an e
 xisting algorithm. Based on this method\, I will present a practical algor
 ithm that takes a set of gene tree topologies and infers species trees wit
 h maximum likelihood. In this algorithm\, we search for the best species t
 ree by starting from candidate species trees found by a parsimony method a
 nd performing local search to obtain better trees with higher likelihood. 
 This algorithm\, called {STELLS}\, has been imp lemented in a program that
  is downloadable from the author's web page. The simulation results show t
 hat the STELLS algorithm is more accurate than several existing methods fo
 r many datasets\, especially when there is noise (in terms of topology\, b
 ranch lengths and rooting) in gene trees. We also show the STELLS algorith
 m is efficient and can be applied to real biological datasets.\n
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR