Paper Title
A Boolean Model on Identification of Signaling Pathways through Supervised and Unsupervised Methods

Abstract
In this study, we focus on applying discrete modeling on ER-positive/ HER2-negative breast cancer cells to identify the main cellular states, caused by mutations in four main signaling pathways. Our main goal lies in identifying these signaling pathways by applying Boolean modeling on gene regulatory networks. For this reason, we use a curated regulatory network of genes, transcription factors, and receptors that characterize the cell profile and then apply a Boolean model on the regulatory network to study the system’s asymptotic behavior, when starting from all the possible initial conditions. Our simulation results reproduce the main cellular states (phenotypes), which were matched to the system attractors first by applying literature-based constraints and subsequently using unsupervised clustering of the attractors according to their binary similarities. Importantly, the results show that the unsupervised method can successfully reproduce the cell states, even in the absence of prior knowledge. Moreover, the model might be used to perform simulations mimicking different mutant conditions that can disrupt specific pathways during tumor progression. Keywords - Gene Regulatory Network, Luminal Breast Cancer, Boolean Model, Signaling Pathways, Jaccard-Needham Distance