Subterranean Fishes of the World

5202

2017

Scientific Data

1-13

4:160132

Fueled by recent advances in comparative genomics, teleost fishes are becoming increasingly important research objects in several scientific disciplines, ranging from ecology, physiology and evolution to medicine, cancer research and aquaculture1–7. Genome information from non-model organisms is highly important in these comparative genomic analyses as they represent specific phenotypes that aid in disentangling the common parts of gene sets from those that have evolved as adaptations to specific ecosystems. In a quest to identify the evolutionary origin of the MHC II pathway loss first observed in the Atlantic cod (Gadus morhua)8,9, we applied a single sequencing library procedure to cost-efficiently produce draft assemblies for 66 teleost species, representing all major lineages within teleost fishes10. Since the alternative immune system, characterized by both the lack of MHC II and an expansion of MHC I, has so far only been identified in the Atlantic cod, we sampled the cod-like fishes of the order Gadiformes more densely than other groups, including 27 species of this order. Based on these genome sequence data, we were able to reconstruct the evolutionary history of the sampled lineages, to pinpoint the loss of the MHC II pathway to the common ancestor of all Gadiformes, and to identify several independent expansions in MHC I copy number within and outside the order Gadiformes. While these analyses and results are reported in a companion paper (Malmstrøm et al.11), we here present in greater detail the underlying data sets used for these analyses, including samples, sequencing reads (Data Citation 1), draft assemblies (Data Citation 2), and both mitochondrial and nuclear phylogenetic markers. By providing these data and the applied methodology in a coherent manner we aim to supply the scientific community with a highly diverse, reliable, and easy-to-use genomic resource for future comparative studies. Our sequencing strategy was chosen on the basis of several pseudo-replicates of the budgerigar (Melopsittacus undulatus) genome12 (Data Citation 3), comprising different combinations of read lengths and coverages to determine the most cost-effective manner to produce genome data of sufficient quality for a reliable determination of gene presence or absence. These budgerigar data sets were furthermore assembled with two of the most used assemblers, the de Bruijn graph based SOAPdenovo13 and the Overlap-Layout-Consensus based Celera Assembler14 to investigate which assembly algorithm performed best on the various data replicates. On the basis of these in silico experiments, all species were sequenced on the Illumina HiSeq2000 platform, aiming for ~15 × coverage. The sequenced reads were then quality controlled, error corrected and trimmed before performing assembly with Celera Assembler. The continuity of the assemblies was subsequently assessed through N50 statistics and the assembly quality was evaluated on the basis of gene space completeness of highly conserved genes. The assemblies were further used to identify mitochondrial genome sequences, which we used in combination with previously available sequences of related teleosts to verify the phylogenetic positions of sampled taxa (Data Citations 4 to 124). By recovering all taxa in their expected positions, clustering with conspecific or congeneric sequences where such were available, our phylogenetic analysis corroborates the correct identification of all sampled taxa and the absence of DNA contamination. Figure 1 illustrates the total workflow, and detailed information for each analysis step is further provided in the Methods section and in Tables 1–7 (available online only). The data sets presented here contain sequencing reads and assembled draft genomes for non-model species adapted to a wide variety of habitats, ranging from the deep sea and tropical coral reefs, to rivers and freshwater lakes. These data sets can be used individually or collectively, as resources for studies such as gene family evolution, adaptation to different habitats, phylogenetic inference of teleost orders, transposons and repeat content evolution as well as many other applications regarding gene and genome evolution in a comparative or model organism framework.

1