Computational Approaches to Study the Relation Between Genomic Variations and Phenotypes

Аннотация

Understanding how individual proteins are organized into complexes and pathways is a significant current challenge. We introduce new algorithms to infer protein complexes by combining seed proteins with a confidence-weighted network. Two new stochastic methods use averaging over a probabilistic ensemble of networks, and a new deterministic method provides a fast ranking of prospective complex members. We compare the performance of these algorithms with three existing algorithms. We test algorithm performance using three methods to assign confidence weights to interaction graphs: a Naive Bayes estimate of the probability of a direct and stable protein-protein interaction; a logistic regression estimate of the probability of a direct or indirect interaction; and a decision tree estimate of whether two proteins exist within a common protein complex. The best-performing algorithms in these trials are the new stochastic methods. The deterministic algorithm is substantially faster, whereas the stochastic algorithms are less sensitive to the weighting scheme.