Domainoid

Domainoid infers orthology at the protein-domain level to identify evolutionary relationships and discordant domain histories that full-length sequence methods can miss.

Key Features:

• Domain-level orthology inference: Infers orthologous relationships for individual protein domains rather than full-length proteins to detect discordant domain histories.
• Adaptation of the InParanoid algorithm: Applies the InParanoid algorithm to individual domains to identify and group orthologous domains across species.
• Protein-level orthology aggregation: Infers protein-level orthology by aggregating domain-level results based on the proportion of domains that are orthologous.
• Benchmark validation: Domain-based assignments have been validated against conventional full-length sequence approaches such as InParanoid using standard benchmarks.

Scientific Applications:

• Functional genomics: Provides domain-resolved evolutionary histories to interpret functional implications of domain architecture variations.
• Evolutionary biology: Detects discordant domain orthologs to study protein evolution driven by recombination and other genetic events.
• Comparative genomics: Enables comparison of domain architectures across species to assess evolutionary conservation and divergence at the domain level.

Methodology:

Protein sequences are dissected into constituent domains and the InParanoid algorithm is applied independently to each domain; protein-level orthology is inferred by aggregating domain-level orthology according to the proportion of orthologous domains.