Abstract
Proteolysis-Targeting Chimeric Molecules (PROTAC) is a rapidly emerging technology for drug target protein degradation and drugging undruggable drug targets. There is a growing literature on in-silico approaches to the complex problem of PROTAC design, specifically the advantages of AI/ in-silico methods in the PROTAC Design Make Test Analyze (DMTA) cycle. Our work presented here aims to contribute to the growing literature of in-silico approaches to PROTAC design by incorporating and demonstrating incremental advancement over previously published methods. We use AI based generative methods for PRTOAC design and Molecular Dynamics to evaluate the stability of the ternary complex formed and ability of the PROTAC to hold the target protein and E3 ligase together stably. To quantify the performance of the PROTAC candidate, we also estimate computationally the PROTAC performance metrics routinely measured by the experimentalists in PROTAC assays. We use highly accurate absolute binding free energy calculations used traditionally in protein-ligand space for the PROTAC system. We calculate (Gibbs free energy change) ΔG for binary complex formation and ternary complex formation mediated by the PROTAC using Free Energy Perturbation - Thermodynamics Integration (FEP-TI) method which is benchmarked in literature with a root mean square error of 0.8 kcal/mol. We calculate ΔG for ternary and binary complexes and estimate whether ΔG for ternary is lower than the ΔG estimated for binary complexes. When the ΔG for ternary is lower than the binary it is inferred that ternary complexation is favoured over binary. Therefore, through these methods we can theoretical estimate ΔG measured by experimentalists in PROTAC assays such as Isothermal titration calorimetry (ITC) and Surface plasmon resonance (SPR) which capture the ΔG for ternary and binary complex formation mediated by the PROTAC. This method will help reduce time as well as costs of the PROTAC DMTA cycle and will accelerate early stage PROTAC drug discovery. As an illustrative application of our in-silico PROTAC design approach, we chose the target Fibroblast growth factor receptor 1(FGFR-1) which is a target approved drug for colorectal cancer. We report the findings and conclude with future research directions.
Competing Interest Statement
The authors have declared no competing interest.