Reactivities of acrylamide warheads toward cysteine targets: a QM/ML approach to covalent inhibitor design.

Covalent inhibition offers many advantages over non-covalent inhibition, but covalent warhead reactivity must be carefully balanced to maintain potency while avoiding unwanted side effects. While warhead reactivities are commonly measured with assays, a computational model to predict warhead reactivities could be useful for several aspects of the covalent inhibitor design process. Studies have shown correlations between covalent warhead reactivities and quantum mechanic (QM) properties that describe important aspects of the covalent reaction mechanism. However, the models from these studies are often linear regression equations and can have limitations associated with their usage. Applications of machine learning (ML) models to predict covalent warhead reactivities with QM descriptors are not extensively seen in the literature. This study uses QM descriptors, calculated at different levels of theory, to train ML models to predict reactivities of covalent acrylamide warheads. The QM/ML models are compared with linear regression models built upon the same QM descriptors and with ML models trained on structure-based features like Morgan fingerprints and RDKit descriptors. Experiments show that the QM/ML models outperform the linear regression models and the structure-based ML models, and literature test sets demonstrate the power of the QM/ML models to predict reactivities of unseen acrylamide warhead scaffolds. Ultimately, these QM/ML models are effective, computationally feasible tools that can expedite the design of new covalent inhibitors.

Dissecting the Interaction Fingerprints and Binding Affinity of BYL719 Analogs Targeting PI3Kα.

Phosphatidylinositol-3-kinase Alpha (PI3Kα) is a lipid kinase which regulates signaling pathways involved in cell proliferation. Dysregulation of these pathways promotes several human cancers, pushing for the development of anticancer drugs to target PI3Kα. One such medicinal chemistry campaign at Novartis led to the discovery of BYL719 (Piqray, Alpelicib), a PI3Kα inhibitor approved by the FDA in 2019 for treatment of HR+/HER2-advanced breast cancer with a PIK3CA mutation. Structure-based drug design played a key role in compound design and optimization throughout the discovery process. However, further characterization of potency drivers via structural dynamics and energetic analyses can be advantageous for ensuing PI3Kα programs. Here, our goal is to employ various in-silico techniques, including molecular simulations and machine learning, to characterize 14 ligands from the BYL719 analogs and predict their binding affinities. The structural insights from molecular simulations suggest that although the ligand-hinge interaction is the primary driver of ligand stability at the pocket, the R group positioning at C2 or C6 of pyridine/pyrimidine also plays a major role. Binding affinities predicted via thermodynamic integration (TI) are in good agreement with previously reported IC50s. Yet, computationally demanding techniques such as TI might not always be the most efficient approach for affinity prediction, as in our case study, fast high-throughput techniques were capable of classifying compounds as active or inactive, and one docking approach showed accuracy comparable to TI.

Predicting In Vivo Compound Brain Penetration Using Multi-task Graph Neural Networks.

Assessing whether compounds penetrate the brain can become critical in drug discovery, either to prevent adverse events or to reach the biological target. Generally, pre-clinical in vivo studies measuring the ratio of brain and blood concentrations (Kp) are required to estimate the brain penetration potential of a new drug entity. In this work, we developed machine learning models to predict in vivo compound brain penetration (as LogKp) from chemical structure. Our results show the benefit of including in vitro experimental data as auxiliary tasks in multi-task graph neural network (MT-GNN) models. MT-GNNs outperformed single-task (ST) models solely trained on in vivo brain penetration data. The best-performing MT-GNN regression model achieved a coefficient of determination of 0.42 and a mean absolute error of 0.39 (2.5-fold) on a prospective validation set and outperformed all tested ST models. To facilitate decision-making, compounds were classified into brain-penetrant or non-penetrant, achieving a Matthew's correlation coefficient of 0.66. Taken together, our findings indicate that the inclusion of in vitro assay data as MT-GNN auxiliary tasks improves in vivo brain penetration predictions and prospective compound prioritization.