Inferring phylogenetic trees from intramolecular distances with Phylo-IMD

Autor: Mansouri, Leila, Baltzis, Athanasios, Langer, Björn E, Magis, Cedrik, de Vienne, Damien M, Gascuel, Olivier, Notredame, Cedric
Jazyk: angličtina
Rok vydání: 2022
Předmět:
DOI: 10.5281/zenodo.7437268
Popis: Phylogenetic reconstruction is one of the most widely used modelling techniques in biology. While most available methods involve the use of sequence-based comparison it has long been discussed whether structural information could be used to further inform tree reconstruction. In this study, we explored the usability of intra molecular distances as a potential evolutionary character. Using multiple sequence alignment to identify homologous internal distances, we have defined a pairwise comparison metric, IMD, that can be used in combination with any distance-based phylogenetic method and lends itself to the measure of branch support values using a procedure similar to the Felsenstein bootstrap. The IMD metric was validated on a dataset of 508 protein families featuring a total of 10,668 protein sequences with known X-Ray structures. The benchmark was carried out using sequence-based maximum likelihood trees as references. Our most important finding is the observation that the IMD bootstrap support is a strong indicator of topological agreement between the structure-based IMD trees and their sequence-based maximum likelihood counterparts. We also developed a titration procedure that allowed us to determine that all things being equal, the evolutionary signal measured by IMD is roughly twice denser than the equivalent sequence-based signal. Altogether our results suggest the suitability of structure-based phylogeny for evolutionary analysis, but above all, they indicate structure-based phylogeny as a powerful means of providing new supporting evidence to protein phylogenetic tree reconstruction. 
Databáze: OpenAIRE