MMPdb and MitoPredictor: Tools for facilitating comparative analysis of animal mitochondrial proteomes

Abstract

Data on experimentally-characterized animal mitochondrial proteomes (mt-proteomes) are limited to a few model organisms and are scattered across multiple databases, impeding a comparative analysis. We developed two resources to address these problems. First, we re-analyzed proteomic data from six species with experimentally characterized mt-proteomes: animals (Homo sapiens, Mus musculus, Caenorhabditis elegans, and Drosophila melanogaster), and outgroups (Acanthamoeba castellanii and Saccharomyces cerevisiae) and created the Metazoan Mitochondrial Proteome Database (MMPdb) to host the results. Second, we developed a novel pipeline, "MitoPredictor" that uses a Random Forest classifier to infer mitochondrial localization of proteins based on orthology, mitochondrial targeting signal prediction, and protein domain analyses. Both tools generate an R Shiny applet that can be used to visualize and interact with the results and can be used on a personal computer. MMPdb is also available online at https://mmpdb.eeob.iastate.edu/.

Keywords: Database; Machine learning; Mitochondria; Proteome; Random Forest.