SwiftOrtho

SwiftOrtho performs orthology classification across multiple genomes to identify orthologs, paralogs, and co-orthologs for comparative genomics, phylogenetics, and protein function annotation.

Key Features:

• Orthology classification: Identifies orthologs, paralogs, and co-orthologs across multiple genomes using graph-based relationships.
• Speed and memory efficiency: Optimized for rapid processing and reduced memory usage, enabling analyses on large datasets and on machines with limited RAM (e.g., ~4 GB).
• Graph-based framework: Represents complex orthology relationships as graphs to manage and analyze inter-genome connections.
• Long k-mers for homology search: Employs long k-mers to accelerate sequence similarity searches during homology detection.
• Reduced amino acid alphabet and spaced seeds: Uses a reduced amino acid alphabet and spaced seeds to mitigate sensitivity loss from long k-mers and maintain homology detection accuracy.
• Affinity propagation clustering: Applies an affinity propagation algorithm to cluster large-scale orthology relationships into orthologous groups while reducing memory usage during clustering.

Scientific Applications:

• Comparative genomics: Enables large-scale comparison of genetic content across species, including analyses of thousands of genomes.
• Phylogenetics: Supports inference of evolutionary relationships by providing orthology and paralogy assignments for downstream phylogenetic analyses.
• Protein function annotation: Facilitates transfer and prediction of protein function based on identified homologs and orthologous groups.

Methodology:

Uses long k-mers for homology search, a reduced amino acid alphabet and spaced seeds to preserve sensitivity, a graph-based representation of relationships, and affinity propagation to cluster sequences into orthologous groups.