Meeting Abstract

65.2  Friday, Jan. 6  A 500-locus phylogenomic study of ray-finned fishes ALFARO, ME*; FAIRCLOTH, B; SORENSON, L; CHANG, J; SANTINI, F; UCLA; UCLA; UCLA; UCLA; UCLA michaelalfaro@ucla.edu

Genomic approaches have helped resolve some of higher-level relationships on the fish tree of life. One key challenge that remains is to create phylogenomic techniques that easily, efficiently, and universally generate large data sets from fishes. Massively parallel sequencing (MPS) offers tremendous potential in this area, yet it has proven difficult to efficiently scale DNA inputs to the output of MPS platforms. For example, the amplification of hundreds of orthologous gene regions across deeply diverging taxa remains an inefficient step compared to sequencing these amplicons in multiplex. To overcome these barriers we developed a technique to universally collect data from over 500 loci across bony fishes.  The method uses sequence capture probes to enrich organismal DNA for hundreds to thousands of ultra-conserved nuclear DNA regions and the sequence flanking these regions.  Following enrichment, we use MPS to sequence captured loci and programmatic tools to align, analyze, and generate species trees from these sequence data. We tested this approach by enriching loci from DNA libraries prepared from 18 species of fish distributed across bony fishes.  We sequenced enriched DNA using an Illumina GAIIx, and we developed a bioinformatic pipeline to assemble, align, and integrate these data with extant genome sequence data from seven fishes prior to phylogenetic analysis. From these alignments, we generated a concatenated matrix composed of 25 species and 160,000 characters.  Bayesian analysis (MrBayes 3.1) of the data reveals a robustly supported tree (posterior probabilities for all nodes >0.99) that is largely consistent with studies based upon other molecular data sets while providing new evidence for the position of the elopomorphs as sister to other teleosts. This highlights the power of a very large number of UCE regions to resolve phylogenetic relationships at both broad and shallow scales.