TY - JOUR T1 - Modeling heterogeneous populations using Boolean networks JF - bioRxiv DO - 10.1101/110817 SP - 110817 AU - Brian C. Ross AU - Mayla Boguslav AU - Holly Weeks AU - James Costello Y1 - 2017/01/01 UR - http://biorxiv.org/content/early/2017/02/22/110817.abstract N2 - Boolean networks are commonly used to model biological pathways and processes, in part because analyses can often find all possible long-term outcomes. Here we describe a Boolean network analysis that captures both the long-term outcomes of a heterogeneous population, as well as the transient behavior leading up to those outcomes. In contrast to other approaches, our method gives an explicit simulation through time using the composition of the mixed population without having to track each subpopulation individually, thus allowing us to simulate heterogeneous populations. This technique accurately models the dynamics of large populations of deterministic, probabilistic or continuous-time Boolean networks that use either synchronous or asynchronous updating. Our method works by treating the network dynamics as a linear system in a variable space that includes products of the Boolean state variables. We show that these product-basis analyses can help find very rare subpopulations or behaviors that sampling-based analyses would likely miss. Such rare events are critical in processes such as the initiation and progression of cancer, and the development of treatment resistance ER -