Integrating Sequence Evolution into Probabilistic Orthology Analysis.

Area of Science:

• Computational Biology
• Genomics
• Phylogenetics

Background:

• Orthology analysis is crucial for understanding gene function and evolution.
• Current methods often infer gene trees independently of species trees, limiting accuracy.
• Existing approaches may not fully leverage sequence data or reconciliation possibilities.

Purpose of the Study:

• To introduce DLRSOrthology, a novel Bayesian Markov chain Monte Carlo (MCMC) method for computing orthology probabilities.
• To improve upon existing orthology inference methods by integrating gene trees, species trees, and sequence data more effectively.
• To assess the performance of DLRSOrthology against established probabilistic orthology approaches.

Main Methods:

• Developed a Bayesian MCMC framework (DLRSOrthology) for orthology probability computation.
• DLRSOrthology jointly considers gene trees, species tree reconciliations, and sequence data.
• Compared DLRSOrthology with PrIME-GEM and MrBayesMPR on synthetic and biological datasets.

Main Results:

• DLRSOrthology demonstrated superior performance compared to PrIME-GEM and MrBayesMPR.
• The method showed robustness even with incomplete taxon sampling.
• DLRSOrthology effectively integrates gene tree, species tree, and sequence information.

Conclusions:

• DLRSOrthology provides a more accurate and comprehensive approach to orthology inference.
• The method's ability to sum over gene trees and consider reconciliations enhances reliability.
• DLRSOrthology represents a significant advancement in computational methods for evolutionary biology.