Probabilistic inference of lateral gene transfer events

Area of Science:

• Evolutionary biology
• Molecular evolution
• Genomics

Background:

• Lateral gene transfer (LGT) is a significant evolutionary mechanism challenging traditional tree-like models of species evolution.
• LGT plays a crucial role in the spread of antibiotic resistance, making its study vital in molecular biology.
• Accurately placing LGT events within species phylogenies, especially with gene duplication and loss, remains a complex challenge.

Purpose of the Study:

• To develop and present novel probabilistic methods for modeling and inferring lateral gene transfer events.
• To accurately reconcile gene trees with dated species trees, incorporating gene duplication and loss.
• To estimate the placement of LGT events and identify potential LGT "highways".

Main Methods:

• A probabilistic method employing Markov Chain Monte Carlo (MCMC) sampling of reconciliations between gene and dated species trees.
• Utilizes the DLTRS probabilistic model, integrating LGT, gene duplication, gene loss, and sequence evolution under a relaxed molecular clock.
• Estimates posterior distributions on gene trees and the precise locations of LGT events.

Main Results:

• The proposed method successfully infers true LGT events and reconciles them to correct species tree edges in simulation studies.
• Application to Cyanobacteria and Molicutes gene families identified potential LGT highways.
• The findings corroborate existing research and reveal previously undetected instances of gene transfer.

Conclusions:

• The developed probabilistic method provides accurate inference of LGT events and their placement on species trees.
• The identification of LGT highways offers new insights into gene flow and evolutionary dynamics.
• This approach advances the study of evolutionary processes influenced by lateral gene transfer.