 |
Center for Nonlinear Studies |
Recent advancements in pharmaceutical chemistry have led to the discovery of many molecules with potential therapeutic benefits. However, a significant challenge in drug development is that 90% of these molecules are poorly soluble. One of the promising solutions to this issue is using cocrystals, which are crystalline solids made up of two different compounds (target bioactive molecule + inert component).My graduate research uses a custom-developed 2D model for molecular dynamics (MD) simulations to the study the cocrystallization process in-depth. Using this scalable GPU-enabled model, we can simulate a wide range of molecular shapes and interactions. The variety of outcomes we observe, and the energetic benefits of formation correlate with the experimentally confirmed cocrystals on a qualitative level.Using large-scale simulations, we have conducted the first comprehensive investigation of cases where cocrystallization does not occur (“missesâ€). In contrast to the popular belief, we have found that thermodynamics alone is not predictive as 85% of cocrystals formed between poor single-component crystallizers and 27% between good ones show an energetic preference for cocrystal formation, yet no cocrystallization occurs. This discovery opens relevant questions about predictive driving factors behind cocrystallization. In my talk, I will discuss our recent advances in investigating the robustness of cocrystallization prediction techniques and what is missing in their adequate description.
Host: Kipton Barros