3DFI: a pipeline to infer protein function using structural homology

Abstract

Inferring protein function is an integral part of genome annotation and analysis. This process is usually performed in silico, and most in silico inferences are based on sequence homology approaches, which can fail when in presence of divergent sequences. However, because protein structures and their biological roles are intertwined, protein function can also be inferred by searching for structural homology. Many excellent tools have been released in recent years with regards to protein structure prediction, structural homology searches and protein visualization. Unfortunately, these tools are disconnected from each other and often use a web server-based approach that is ill-suited to high-throughput genome-wide analyses. To help assist genome annotation, we built a structural homology-based pipeline called 3DFI (for tridimensional functional inference) leveraging some of the best structural homology tools. This pipeline was built with simplicity of use in mind and enables genome-wide structural homology inferences.