Additional data to: Deep metazoan phylogeny: When different genes tell different stories

Abstract

Molecular phylogenetics resulted in a plethora of controversial hypotheses about the early diversification of non-bilaterian animals. To date, increasing the amount of DNA sequence data analyzed has been insufficient to resolve these relationships unequivocally. To unravel the causes for the patterns of extreme inconsistencies at the base of the metazoan tree of life, we constructed a novel supermatrix containing 122 genes, enriched with new non-bilaterian taxa. Comparative analyses of two non-overlapping multi-gene partitions of this dataset revealed conflicting phylogenetic signals. We show that gene sampling correlates with levels of saturation and Long Branch Attraction artifacts in the two partitions. Additional systematic errors are derived from significant variation in amino acid substitution patterns among metazoan lineages that violate the stationary assumption of models frequently used to reconstruct phylogenies. By modifying gene sampling and taxonomic composition of the outgroup, we were able to construct three different yet well-supported phylogenies. These results show that the accuracy of phylogenetic inference may be improved substantially by selecting genes that evolve slowly across metazoan lineages and applying more realistic substitution models. Additional non sequence-based markers are also necessary to assess the validity of phylogenetic reconstructions.