HLRB Project h0663
Disentangling the Phylogeny of Mammals: Large-scale Multi-Gene Analyses with Maximum Likelihood
LRR, Institut für Informatik, TU München
Proposing Institution
LRR, Institut für Informatik, TU München
Project Manager
Michael Ott
Boltzmannstr. 3
85748 Garching b. München
Abstract
Phylogenetic inference represents a grand challenge for Bioinformaticsdue to immense computational requirements. The increasing popularityof multi-gene alignments in biological studies, which typically provide a more stable topological signal due to a more favorable ratio of the number of base pairs to the number of sequences, coupled with rapid accumulation of sequence data in general, poses new challenges for highperformance computing. Recent advances in the adaptation of MaximumLikelihood (ML) based tree inference programs to massively parallel machines in conjunction with novel algorithmic methods for computationof non–parametric Bootstrap (BS) confidence values on trees allow for analysis of such challenging datasets. Within the framework of the proposed project we intend to adapt RAxML, which is currently the fastest available ML-based inference program to the HLRB-II. The goals of the project are two-fold: Firstly, to adapt RAxML to the HLRB-II architecture, integrate the novel BS method into the parallel code, and make it available to the broader user community. Secondly, to conduct a large-scale study of mammalian evolutionary relationships based on one of the taxonomically most comprehensive multi-gene data sets currently available.
