Advances in computer hardware, the availability of HPC resources, along with artificial intelligence (AI) methods are successfully complementing traditional approaches in medicinal chemistry. In particular, machine learning is gaining importance with the growth of available data collections. Since the outbreak of the SARS-CoV-2 pandemic in early 2020, tremendous scientific efforts have been made in the field of novel antiviral (or coronaviral) compounds. Immediately in 2021, new variants of the virus with several key mutations emerged, with variants of concern named Delta (B.1.617.2 and AY lineage) and Omicron (B.1.1.529 and BA lineages). In this context, we immediately targeted 3CLpro, the main viral protease, as a valuable therapeutic target that possesses no mutations at the active site and is a viable pan-coronaviral target. Therefore, we performed a high throughput virtual screening experiment using CmDock software and a commercially available chemical library to prepare prioritisation lists of compounds for further study. We coupled the classical HTVS approach with a machine learning assisted classification and activity regression study to investigate whether the current ChEMBL-based SARS-CoV-2 activity data could be effectively used and the specificities of the chemical space under investigation could be identified. Indeed, we successfully classified the HTVS screening hits according to 3CLpro inhibitors, viral cysteine protease inhibitors, or the rest of the chemical space based on the calculated set of 208 chemical descriptors. Last but not least, we analysed whether the current set of 3CLpro inhibitors can be used for activity prediction and found that the 3CLpro inhibitor space is underrepresented and not suitable for such a regression task. We postulate that this methodology of 3CLpro inhibitor library preparation and compound prioritisation far outperforms the selection of compounds from available commercial "corona-focused libraries" and boosts traditional HTVS approaches.

Prioritisation of compounds for 3CLpro inhibitor identification using machine learning supported HTVS

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.