Synteny identifies reliable orthologs for phylogenomics and comparative genomics of the Brassicaceae

Abstract

AbstractLarge genomic datasets are becoming the new normal for any facet of phylogenetic research, but identification of true orthologous genes and exclusion of problematic paralogs is still challenging when applying commonly used sequencing methods such as target enrichment. Here, we compared conventional ortholog detection using OrthoFinder with ortholog detection through genomic synteny in a dataset of eleven representative diploid Brassicaceae whole genome sequences spanning the entire phylogenetic space. We then evaluated the resulting gene sets regarding gene number, functional annotation, gene and species tree resolution. Finally, we used the syntenic gene sets for comparative genomics and ancestral genome analysis. The use of synteny resulted in considerably more orthologs and also allowed us to reliably identify paralogs. Surprisingly, we did not detect notable differences between species trees reconstructed from syntenic orthologs compared other gene sets, including the Angiosperm353 set and a Brassicaceae specific target enrichment gene set. However, the synteny dataset comprised a multitude of gene functions, strongly suggesting that this method of marker selection for phylogenomics is suitable for studies that value downstream gene function analysis, gene interaction and network studies. Finally, we present the first ancestral genome reconstruction for the Core Brassicaceae predating the Brassicaceae lineage diversification ∼25 million years ago.