Coordinated Gene Activity of Pattern Sets (CoGAPS)

Abstract

Cancer, one of the leading causes of death, is a group of complex genetic diseases that result from changes to genes which cause abnormal cell growth. CoGAPS, an R package that utilizes a Bayesian Markov Chain Monte Carlo matrix factorization algorithm, can be used to infer biological process activity. By improving the use of data and prior information, the power of the algorithm to identify biologically meaningful patterns will be greatly enhanced. This will help further a systems-level understanding of cell behavior and improve biological models of cell interactions. This information can be used in conjunction with signaling network models to identify active biological processes and the effects of targeted treatments. Thus, it will become easier to more accurately measure how well a cancer treatment is performing and if it has any unintended side-effects, to better understand, diagnose, and treat cancer, and to decrease the mortality rate of those afflicted with cancer.