**GTA** is a cytoscape module that given a network and normal/case gene expressions, finds candidate subnetworks related to cases. For example, GTA given a network and gene expressions related to cancer, finds candidates subnetworks related to cancer.
The theory behind GTA is that, for complex diseases like cancer, single genes are not appropriate markers. On the other hand, GTA suggest subnetworks as cancer biomarkers.
GTA works based on game theory. It tries to find modules which are Nash equilibrium for all participating genes, based on some payoff functions.
For more information and technical details you may refer to the manuscript that proposed this algorithm on [http://pubs.rsc.org/en/Content/ArticleLanding/2016/MB/C5MB00684H#!divAbstract Farahmand, S., et al. "GTA: a game theoretic approach to identifying cancer subnetwork markers." Molecular BioSystems 12.3 (2016): 818-825.]