JustOrthologs: a fast, accurate and user-friendly ortholog identification algorithm

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Orthologous gene identification is crucial for understanding gene function, evolutionary relationships, and phylogenetic inference.
• Current ortholog identification methods, often relying on all-versus-all BLAST comparisons, are computationally intensive and time-consuming.
• There is a need for faster and more efficient algorithms for large-scale ortholog discovery.

Purpose of the Study:

• To develop and present a novel, rapid algorithm for orthologous gene identification.
• To significantly reduce the computational resources required for ortholog detection.
• To enable large-scale comparative genomics studies through accelerated ortholog identification.

Main Methods:

• The JustOrthologs algorithm leverages conserved gene structure, specifically coding sequence lengths and dinucleotide percentages, for ortholog identification.
• It avoids computationally expensive all-versus-all BLAST searches.
• The method was compared against established tools like OrthoMCL, OMA, and OrthoFinder.

Main Results:

• JustOrthologs achieved a runtime reduction of over 96% compared to existing methods while maintaining comparable precision and recall.
• The speedup enabled pairwise comparisons of 1197 complete genomes (780 eukaryotic, 417 archaeal).
• The analysis confirmed annotations for over 384,000 genes, identified over 1.6 million new ortholog groups, and flagged over 51,000 potentially mislabeled genes.

Conclusions:

• JustOrthologs provides a highly efficient and accurate solution for orthologous gene identification.
• The software facilitates large-scale comparative genomics and aids in improving genome annotation quality.
• This approach significantly advances the speed and scalability of ortholog discovery in biological research.