Genome assembly variation and its implications for gene discovery in nematode species

Abstract

BackgroundReduced sequencing costs is facilitating comprehensive BioGenome projects and sequencing of specific species by individual research groups. While the choice between sequencing technologies is limited, the choice of assembly algorithm can be overwhelming. However, this choice will influence the accuracy of the assembly and downstream analysis and annotation. Here, we set out to demonstrate the extent and effect of assembly variation using three nematode species—Caenorhabditis bovis,Haemonchus contortus,Heligmosomoides bakeri—whose genomes differ in size and complexity. In addition to comparing the assembly accuracy and structure, we predicted gene models from each assembly and investigated the effect on large gene family membership.ResultsWe found that assemblers using the Overlap-Layout-Consensus (OLC) algorithm were better choices for assembling more complex genomes while those that used the De-Bruijn Graph (DBG) algorithm performed better in smaller genomes. However, the DBG Redbean assembler performed comparably well in both scenarios, making it a versatile choice. Using standard assembly metrics, variations seemed relatively minor; however, we found noticeable intra-species/inter-assembly differences in the composition of four gene families—CYP, GST, UGT, and NHR—that are of interest in anti-nematode drug discovery and drug resistance research. Finally, through comparisons of highly conserved genes in the assemblies ofH. contortus, we found probable errors in the highly curated reference assembly, demonstrating the need to support ongoing curation efforts.ConclusionsWhile the choice of assembly algorithm can be daunting and the differences appear relatively minor, there is a need to consider differences in assemblies in terms of the assembly structure and subsequent annotation. We repeat the challenge, started by others, of relying on a small number of convenient assembly metrics. Finally, we demonstrate the value of exploring alternative assemblies, particularly in the study of content of protein-coding genes.