Integrating gene annotation with orthology inference at scale

Abstract

Annotating coding genes and inferring orthologs are two classical challenges in genomics and evolutionary biology that have traditionally been approached separately, limiting scalability. We present TOGA (Tool to infer Orthologs from Genome Alignments), a method that integrates structural gene annotation and orthology inference. TOGA implements a different paradigm to infer orthologous loci, improves ortholog detection and annotation of conserved genes compared with state-of-the-art methods, and handles even highly fragmented assemblies. TOGA scales to hundreds of genomes, which we demonstrate by applying it to 488 placental mammal and 501 bird assemblies, creating the largest comparative gene resources so far. Additionally, TOGA detects gene losses, enables selection screens, and automatically provides a superior measure of mammalian genome quality. TOGA is a powerful and scalable method to annotate and compare genes in the genomic era.

Vocal learning–associated convergent evolution in mammalian proteins and regulatory elements

“…To explore the possibility of shared genomic specializations associated with vocal learning, we first applied RERconverge ( 11 ) to recently released protein-coding alignments obtained for hundreds of mammals ( 10 ) to identify protein-coding genes whose relative rates of evolution differ between vocal learners and other mammals and those that may thus be under selection related to vocal learning ( 11 , 24 ). We analyzed 16,209 high-quality protein-coding gene alignments across 215 boreoeutherian mammals, including 26 vocal learning species, 164 vocal nonlearners, and 25 species without confident annotations (Fig.…”

“…To find vocal learning–associated convergent evolution in protein-coding sequences of the mammalian genome, we began with amino acid–level multiple sequence alignments produced by the Zoonomia consortium ( 10 ). Those served as input to two classes of methods, RERconverge ( 11 ) and HyPhy ( 30 ).…”

“…Neurobiology section, Division of Biological Science, University of California, San Diego, La Jolla, CA 92093, USA 9. Molecular Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037, USA 10. Department of Biological Sciences, University of Pittsburgh, Pittsburgh, PA 15213, USA 11.…”

“…1A)—using whole-genome datasets and recently developed computational approaches. Specifically, we used protein-coding sequences from genomes generated by the Zoonomia Consortium ( 9 , 10 ) and models of evolutionary rate convergence ( 11 , 12 ) to identify protein-coding regions repeatedly associated with the evolution of vocal learning across mammals. Although we found 200 protein-coding genes significantly associated with vocal learning, none of them showed strong evidence of selection in all four mammalian clades, and only five showed strong evidence of selection in three out of the four clades.…”

Interactive visualization of whole eukaryote genome alignments using NCBI’s Comparative Genome Viewer (CGV)

“…The second type of comparison data is whole genome assembly alignments (e.g., Mauve (18), LASTZ (19)), which are sequence-based and include both genic and intergenic regions. Whole genome alignments can be much more complex than simple gene ortholog locations but have the advantage of including alignments in regulatory regions and other regions not annotated as genes.…”