DIGITAL ARCHIPELAGO

‘Bioinformatic selection of models of protein evolution’

Biomedicine and Health

Biomedical Research

2004

Models of amino acid replacement are extensively used to study the evolution of proteins, and therefore are essential to reconstruct protein phylogenies and to study changes in protein function or structure. The use of inappropriate protein models can bias study results, and it is therefore very important to justify the use of one or other model. However researchers using models of protein evolution currently lack a tool for selecting the most appropriate model for their data.

DIRECTOR

David Posada González

RESEARCH TEAM

David Posada González, Rafael Zardoya San Sebastian.

COLLABORATING INSTITUTIONS

Universidad de Vigo

DESCRIPTION

The project sets out to fill this gap through the development of a bioinformatic tool for the selection of best-fit models of amino acid replacement from protein alignments. Using information theory, this tool will also implement diagnostics for the assessment of model selection uncertainty and multimodel inference. A statistical framework will be employed for the selection of models of protein evolution. The tool will be applied to a variety of protein alignments to study the patterns of aminoacid replacement along the mitochondrial genome and across different vertebrate lineages. This will allow us to validate its use and, at the same time, to study changes in functional constraints in those proteins evolving under different replacement schemes. Finally, to guarantee its successful dissemination, we will build a web server from which researchers can download this bioinformatic tool and also analyze their data online.