Small Parsimony for Natural Genomes in the DCJ-Indel Model

TL;DR

This paper introduces an Integer Linear Program method to solve the Small Parsimony Problem for natural genomes under the DCJ-indel model, effectively reconstructing ancestral gene orders with high accuracy.

Contribution

It presents a novel ILP algorithm for ancestral genome reconstruction considering gene duplication, loss, and complex rearrangements in the DCJ-indel model.

Findings

Efficient and accurate reconstruction of ancestral genomes.

Applicable to complex evolutionary models including duplications and indels.

Validated on simulated data with promising results.

Abstract

Reconstructing ancestral gene orders is an important step towards understanding genome evolution. The Small Parsimony Problem (SPP) has been extensively studied in this regard. The problem aims at finding the gene orders at internal nodes of a given phylogenetic tree such that the overall genome rearrangement distance along the tree branches is minimized. However, this problem is intractable in most genome rearrangement models, especially when gene duplication and loss are considered. In this work, we describe an Integer Linear Program algorithm to solve the SPP for natural genomes, i.e., genomes that contain conserved, unique, and duplicated markers. The evolutionary model that we consider is the DCJ-indel model that includes the Double-Cut and Join rearrangement operation and the insertion and deletion of genome segments. We evaluate our algorithm on simulated data and show that it is able to reconstruct very efficiently and accurately ancestral gene orders in a very comprehensive evolutionary model.