A contradiction-based framework for testing gene regulation hypotheses

Abstract

We have developed a mathematical framework for representing and testing hypotheses about gene, protein, and signaling molecule interactions. It takes a hierarchical, contradiction-based approach, and can make use of multiple data sources to assess hypothesis viability and to generate a viability partial order over the space of hypotheses. We have developed an event-based formal language for the expression of such hypotheses. This language seamlessly integrates regulatory diagrams (graphical inputs) and structured English (text input) to maximize flexibility. We have developed a pre-topological formalism that allows us to make precise statements about hypothesis similarity and the convergence of iterative refinements of a base hypothesis. To this, we add mathematical machinery that allows us to make precise statements about control and regulation.