Support Vector Machine as a Supervised Learning for the Prioritization of Novel Potential SARS-CoV-2 Main Protease Inhibitors.

In the last year, the COVID-19 pandemic has highly affected the lifestyle of the world population, encouraging the scientific community towards a great effort on studying the infection molecular mechanisms. Several vaccine formulations are nowadays available and helping to reach immunity. Nevertheless, there is a growing interest towards the development of novel anti-covid drugs. In this scenario, the main protease (Mpro) represents an appealing target, being the enzyme responsible for the cleavage of polypeptides during the viral genome transcription. With the aim of sharing new insights for the design of novel Mpro inhibitors, our research group developed a machine learning approach using the support vector machine (SVM) classification. Starting from a dataset of two million commercially available compounds, the model was able to classify two hundred novel chemo-types as potentially active against the viral protease. The compounds labelled as actives by SVM were next evaluated through consensus docking studies on two PDB structures and their binding mode was compared to well-known protease inhibitors. The best five compounds selected by consensus docking were then submitted to molecular dynamics to deepen binding interactions stability. Of note, the compounds selected via SVM retrieved all the most important interactions known in the literature.

Key genetic elements, single and in clusters, underlying geographically dependent SARS-CoV-2 genetic adaptation and their impact on binding affinity for drugs and immune control.

OBJECTIVES: To define key genetic elements, single or in clusters, underlying SARS-CoV-2 (severe acute respiratory syndrome coronavirus-2) evolutionary diversification across continents, and their impact on drug-binding affinity and viral antigenicity. METHODS: A total of 12 150 SARS-CoV-2 sequences (publicly available) from 69 countries were analysed. Mutational clusters were assessed by hierarchical clustering. Structure-based virtual screening (SBVS) was used to select the best inhibitors of 3-chymotrypsin-like protease (3CL-Pr) and RNA-dependent RNA polymerase (RdRp) among the FDA-approved drugs and to evaluate the impact of mutations on binding affinity of these drugs. The impact of mutations on epitope recognition was predicted following Grifoni et al. (Cell Host Microbe 2020. 27: 671-80.). RESULTS: Thirty-five key mutations were identified (prevalence: ≥0.5%), residing in different viral proteins. Sixteen out of 35 formed tight clusters involving multiple SARS-CoV-2 proteins, highlighting intergenic co-evolution. Some clusters (including D614GSpike + P323LRdRp + R203KN + G204RN) occurred in all continents, while others showed a geographically restricted circulation (T1198KPL-Pr + P13LN + A97VRdRp in Asia, L84SORF-8 + S197LN in Europe, Y541CHel + H504CHel + L84SORF-8 in America and Oceania). SBVS identified 20 best RdRp inhibitors and 21 best 3CL-Pr inhibitors belonging to different drug classes. Notably, mutations in RdRp or 3CL-Pr modulate, positively or negatively, the binding affinity of these drugs. Among them, P323LRdRp (prevalence: 61.9%) reduced the binding affinity of specific compounds including remdesivir while it increased the binding affinity of the purine analogues penciclovir and tenofovir, suggesting potential hypersusceptibility. Finally, specific mutations (including Y541CHel + H504CHel) strongly hampered recognition of Class I/II epitopes, while D614GSpike profoundly altered the structural stability of a recently identified B cell epitope target of neutralizing antibodies (amino acids 592-620). CONCLUSIONS: Key genetic elements reflect geographically dependent SARS-CoV-2 genetic adaptation, and may play a potential role in modulating drug susceptibility and hampering viral antigenicity. Thus, a close monitoring of SARS-CoV-2 mutational patterns is crucial to ensure the effectiveness of treatments and vaccines worldwide.