Proteinortho6: pseudo-reciprocal best alignment heuristic for graph-based detection of (co-)orthologs

• DOI: 10.3389/fbinf.2023.1322477
• Publication Date: 2023-12-13T05:04:59Z
• Authors: Paul Klemm; Peter F. Stadler; Marcus Lechner
• Keywords: Reciprocal

Proteinortho is a widely used tool to predict (co)-orthologous groups of genes for any set species. It finds application in comparative and functional genomics, phylogenomics, evolutionary reconstructions. With rapidly increasing number available genomes, the demand large-scale predictions also growing. In this contribution, we evaluate implement major algorithmic improvements that significantly enhance speed analysis without reducing precision. Graph-based detection (co-)orthologs typically based on reciprocal best alignment heuristic requires an all vs. comparison proteins from species under study. The initial identification similar accelerated by introducing alternative search along with revised strategy-the pseudo-reciprocal heuristic-that reduces required sequence comparisons one-half. clustering algorithm was reworked efficiently decompose very large clusters accelerate processing. Proteinortho6 overall processing time order magnitude compared its predecessor while maintaining small memory footprint good predictive quality.