DOI: 10.1371/JOURNAL.PONE.0062892
关键词: Annelid 、 Prediction methods 、 Genomics 、 Genetics 、 Evolutionary biology 、 Taxon 、 Phylogenetics 、 Biology 、 Clade 、 Supermatrix 、 Expressed sequence tag
摘要: Phylogenomic studies based on hundreds of genes derived from expressed sequence tags libraries are increasingly used to reveal the phylogeny taxa. A prerequisite for these is assignment into clusters orthologous sequences. Sophisticated methods orthology prediction in such analyses, but it rarely assessed whether paralogous sequences have been erroneously grouped together as after prediction, and this had an impact phylogenetic reconstruction using a super-matrix approach. Herein, I tested annelid relationships phylogenomic datasets. Using single-partition screening bootstrap support, blast searches pruning supermatrix, wrongly assigned were found eight partitions placement five taxa (the annelids Owenia, Scoloplos, Sthenelais Eurythoe nemertean Cerebratulus) including robust support could be attributed presence two partitions. Excluding resulted different, weaker supported Moreover, analyses revealed that impacted when only single taxon represented previously higher polychaete family. One possibility priori detection combine 1) criteria nodal or internal branch length with 2) suspicious cases presented herein. Also possible posteriori approaches which specific clades investigated by comparing alternative hypotheses differences per-site likelihoods. Increasing sizes EST will also decrease likelihood sequences, case like HaMStR likewise decreased more than one reference taxon.