UC Santa Barbara

Modeling of biochemical polymers using classical simulation and the analysis of large datasets has become important in an age when high-throughput experimentation and advanced computational resources have enabled the collection of massive bodies of information regarding these systems. Models help us understand and extract insight from experimental data, allowing us to develop simulations and predictions around their behavior that can help elucidate macroscopic behaviors of interest. The body of work applies techniques in molecular dynamics, machine intelligence and information theory to problems within the realm of biological heteropolymers with the intent of developing understanding regarding modern methods of data generation and analysis. The differing paradigms of first-principles models versus data first models are considered and examined, and work is done to show areas where the methods are complementary or applicable to helping understand experimental data sets.