A likelihood based clustering method for detection of recombination for DNA and Amino Acid sequences

Abstract

Genetic Recombination is a process where parts of different genes are combined to form a new gene. Recombination detection is an important part of phylogenetic analysis for DNA sequences and thus the detection of recombination events has received great attention in the phylogenetics literature. However most methods are either computationally expensive and are not suitable for analyzing many genes or a genome, or are computationally fast but are not accurate enough. Furthermore, almost all existing packages are developed for DNA sequences, and cannot be easily used to analyze amino acid sequences. We propose a new algorithm which is fast and accurate for recombination detection and can be used on both DNA or amino acid sequences. Our method is a simple clustering algorithm based on the site log-likelihood. Performance of the method is evaluated on both simulated and real data examples.