Developing effective therapies for myocardial fibrosis is challenging because the fibrotic signaling pathway involves a complex interplay between biomechanical and biochemical stimuli. This complexity is further enhanced by individual patient variability. To address this challenge, we are advancing a large-scale, ODE-based signaling network model (SNM) that contains many well-established fibrotic pathways. Using patient-specific proteomic data as inputs to the SNM can yield simulations of individual biological contexts. In lieu of actual patient data, we first simulated 200 patient-specific contexts by randomizing the model inputs and generating a baseline fibrotic signaling network for each context. We then performed a single perturbation screen on each model, where all nodes were iteratively suppressed to 0.1 times their baseline activation and enhanced to 10 times their baseline activation. This resulted in “patient”-specific drug target rankings. Principal component analysis of the perturbation responses revealed three distinct patient response clusters, suggesting the existence of characteristic input signatures that drive differential fibrotic response. These findings support the feasibility of identifying mechanistically distinct patient subgroups that may benefit from tailored interventions. To investigate this further, we are now simulating biological contexts informed by real-world patient data and extending the analysis to double perturbation screens. Ultimately, this framework enables the prediction of patient-specific drug responses for myocardial fibrosis, potentially guiding the development of precision therapeutic strategies.
