Selecting ribosomal protein genes for invertebrate phylogenetic inferences: How many genes to resolve the Mollusca?

Abstract

1.Multi-gene analyses are currently the gold standard in phylogenetics, despite limited taxon sampling. To facilitate broad taxon representation on an economically tolerable level, we optimize the gene selection for future PCR-based sampling strategies. 2.Highly expressed ribosomal proteins (RP) were sampled chiefly for molluscs, the second largest metazoan phylum with largely unknown internal relationships. Thirty-two new sequences for Lepidochitona cinerea (Polyplacophora) were integrated into a data-matrix of 79 RP genes, comprising 16 mollusc species (five classes). The resulting maximum likelihood tree was used to evaluate each single-gene tree according to its topological fit. Genes were sorted in ascending order, successively concatenated and then evaluated. 3.The most efficient sub-alignment consists of 18 selected genes. All analyses reveal the Mollusca as monophyletic. The phylogenetic utility of the selected and concatenated 18 ribosomal genes is comparable to results from phylogenomic analyses. 4.This optimized data set can be assembled using PCR methods and will allow expanding the taxon coverage in deep molluscan phylogeny far beyond the current sampling. © 2010 British Ecological Society.