Rock, paper, scissors: harnessing complementarity in ortholog detection methods improves comparative genomic...

Area of Science:

• Comparative genomics
• Bioinformatics
• Evolutionary biology

Background:

• Ortholog detection (OD) is crucial for comparative genomics, identifying genes descended from a common ancestor.
• Existing OD methods have varying strengths and weaknesses across different evolutionary and practical scenarios.

Purpose of the Study:

• To compare the performance of four distinct ortholog detection methods across ten mammalian proteomes.
• To develop MOSAIC (Multiple Orthologous Sequence Analysis and Integration by Cluster optimization), a novel tool for integrating diverse OD methods.

Main Methods:

• Proteome analysis of ten mammals using four rigorously filtered, methodologically distinct OD methods.
• Development and application of MOSAIC for integrating results from multiple OD methods.

Main Results:

• Individual OD methods showed significant complementarity, outperforming each other for 38-45% of genes.
• MOSAIC increased the number of complete species alignments by over fivefold while maintaining or improving ortholog quality metrics.
• MOSAIC alignments revealed more confidently aligned sites, higher conservation, and up to 180% more detected positively selected sites.

Conclusions:

• Integrating diverse OD methods via MOSAIC substantially enhances ortholog detection and alignment quality.
• MOSAIC facilitates more robust evolutionary analyses, including the detection of positive selection.
• MOSAIC identified a potential positively selected site in TPSAB1, an enzyme implicated in various diseases.