Premium
Highly parallelized inference of large genome‐based phylogenies
Author(s) -
MeierKolthoff Jan P.,
Auch Alexander F.,
Klenk HansPeter,
Göker Markus
Publication year - 2014
Publication title -
concurrency and computation: practice and experience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.309
H-Index - 67
eISSN - 1532-0634
pISSN - 1532-0626
DOI - 10.1002/cpe.3112
Subject(s) - phylogenetic tree , genome , pairwise comparison , computational biology , biology , inference , phylogenetics , proteome , genomics , phylogenetic network , computer science , genetics , gene , artificial intelligence
SUMMARY Genome Blast Distance Phylogeny ( GBDP ) infers distances and phylogenetic relationships between organisms from completely or partially sequenced genomes. It is well suited for parallelization as pairwise distances are calculated independently. As exemplar data for a high‐performance cluster implementation that executes many pairwise genome comparisons in parallel, we here used sequences from the Genomic Encyclopedia of Bacteria and Archaea project. Phylogenies were inferred from genome‐scale nucleotide and amino acid data with all variants of GBDP , including novel adaptations to amino acid sequences and approaches yielding trees with branch support. The dependency of phylogenetic accuracy, average branch support as well as performance indicators such as running time and disk space consumption on details of genome comparison, distance calculation, and phylogenetic inference was examined in detail. If combined with conservative measures for branch support, GBDP appears to infer reasonable phylogenetic relationships of microorganisms with a comparatively low computational cost. Due to the linear speed‐up of the cluster, benchmarks reveal an overall computation time of less than 24 h required for the 7750 pairwise genome/proteome comparisons of the Genomic Encyclopedia of Bacteria and Archaea data set that is opposed to an estimated running time of about 30 days for the non‐parallelized version. Copyright © 2013 John Wiley & Sons, Ltd.