Modeling transcription factor interaction in biochemically grounded gene regulatory networks

Gene regulatory networks (GRNs) are useful tools to help understand biological pathways on a molecular level. However, it remains challenging to determine the combinatorial effects of transcriptional regulation; understanding how multiple regulators operate and predicting combinatorial regulation is crucial to our understanding of complex systems. Boolean networks have achieved success in using transcription factor (TF) interactions to model combinatorial regulation and classify AND and OR regulation. We seek to go beyond their analysis by modeling gene expression with continuous variables and to identify intermediate states in gene \u27networks. We develop a GRN model that emulates logic gates using the underlying biochemical rate equations, allowing it to show complex TF interaction patterns while remaining biologically significanta nd representingi ntermediatev alues. We then use this simulationa nd machine learning techniques to predict the presence and type combinatorial regulation to a high accuracy in simulated data