Primary orthologs from local sequence context.

Area of Science:

• Genomics
• Evolutionary Biology
• Bioinformatics

Background:

• Inferring gene evolutionary history is vital in biology.
• Conserved non-coding sequences in mammalian genomes require methods beyond protein-coding analysis.
• Distinguishing primary orthologs from other homologs necessitates genomic context, often computationally intensive.

Purpose of the Study:

• To develop a computationally efficient method for identifying primary orthologs using short-range genomic sequence context.
• To overcome limitations of similarity-based and traditional alignment methods in ortholog identification.

Main Methods:

• Utilizing genome intersection to extract "non-nested maximal matches" from mammalian genomes.
• Employing short-range sequence context, as minimal as a single maximal match, for ortholog inference.
• Developing a parameter-free, intersection-based approach without repeat-masking or alignment.

Main Results:

• Non-nested maximal matches accurately identify primary (positional) orthologs with high precision and recall across genomes.
• The method is over 30 times faster than commonly used whole-genome alignment techniques.
• Novel putative orthologs, such as approximately 1000 gene pairs in human-chimpanzee, were identified using reciprocal best hits (mmRBHs).

Conclusions:

• An intersection-based method effectively infers sequence evolutionary history using short-range genomic context.
• This approach is computationally efficient, parameter-free, and suitable for genome-wide ortholog identification.
• The method can identify orthologs in repeat-masked regions and may be applicable to unassembled genomes.