TENET: A Machine Learning-Based System for Target Characterization in Signaling Networks

Abstract

Target characterization of a biological network identifies characteristics that distinguish targets (nodes that can serve as molecular targets of drugs) from other nodes. In this demonstration, we present TENET (Target charactErization using NEtwork Topology), a software that facilitates topological features-based characterization of known targets in signaling networks modelling dynamic interactions within biological systems. TENET is based on a support vector machine (SVM)-based approach and generates a characterization model. These models specify topological features that can discriminate known targets and how these features are combined to quantify the likelihood of a node being a target. Hence, TENET can be used for prioritizing targets and for identifying novel candidate targets that share similar characteristics with known targets. The interactive user interface that TENET provides facilitates users' study and understanding of topological characteristics of targets in signaling networks.