Pre-training molecular representation model with spatial geometry for property prediction.

AI-enhanced bioinformatics and cheminformatics pivots on generating increasingly descriptive and generalized molecular representation. Accurate prediction of molecular properties needs a comprehensive description of molecular geometry. We design a novel Graph Isomorphic Network (GIN) based model integrating a three-level network structure with a dual-level pre-training approach that aligns the characteristics of molecules. In our Spatial Molecular Pre-training (SMPT) Model, the network can learn implicit geometric information in layers from lower to higher according to the dimension. Extensive evaluations against established baseline models validate the enhanced efficacy of SMPT, with notable accomplishments in classification tasks. These results emphasize the importance of spatial geometric information in molecular representation modeling and demonstrate the potential of SMPT as a valuable tool for property prediction.

An Open Source Graph-Based Weighted Cycle Closure Method for Relative Binding Free Energy Calculations.

Free energy perturbation-relative binding free energy (FEP-RBFE) prediction has shown its reliability and accuracy in the prediction of protein-ligand binding affinities, which plays a fundamental role in structure-based drug design. In FEP-RBFE predictions, the calculation of each mutation path is associated with a statistical error, and cycle closure (cc) has proven to be an effective method in improving the calculation accuracy by correcting the hysteresis (summation of errors) of each closed cycle to the theoretical value 0. However, a primary hypothesis was made in the current cycle closure method that the hysteresis is evenly distributed to all paths, which is unlikely to be true in practice and may limit the further improvement of the calculation accuracy when better error estimation methods are available. Moreover, being a closed source software makes the current cycle closure method unachievable in many studies. In this paper, a newly implemented open source graph-based weighted cycle closure (wcc) algorithm was developed and introduced, not only including functions from the original cc method but also containing a new wcc method which can consider different error contributions from different paths and further improve the calculation accuracy. The wcc program also provides a new path-independent molecular error calculation method, which can be quite useful in many studies (like structure-activity relationship (SAR)) compared with the path-dependent method of the original cc program.

Free energy perturbation-based large-scale virtual screening for effective drug discovery against COVID-19.

As a theoretically rigorous and accurate method, FEP-ABFE (Free Energy Perturbation-Absolute Binding Free Energy) calculations showed great potential in drug discovery, but its practical application was difficult due to high computational cost. To rapidly discover antiviral drugs targeting SARS-CoV-2 Mpro and TMPRSS2, we performed FEP-ABFE-based virtual screening for ∼12,000 protein-ligand binding systems on a new generation of Tianhe supercomputer. A task management tool was specifically developed for automating the whole process involving more than 500,000 MD tasks. In further experimental validation, 50 out of 98 tested compounds showed significant inhibitory activity towards Mpro, and one representative inhibitor, dipyridamole, showed remarkable outcomes in subsequent clinical trials. This work not only demonstrates the potential of FEP-ABFE in drug discovery but also provides an excellent starting point for further development of anti-SARS-CoV-2 drugs. Besides, ∼500 TB of data generated in this work will also accelerate the further development of FEP-related methods.

Exploring the Binding Affinity and Mechanism between ACE2 and the Trimers of Delta and Omicron Spike Proteins by Molecular Dynamics Simulation and Bioassay.

Five major variants of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have emerged and posed challenges in controlling the pandemic. Among them, the current dominant variant, viz., Omicron, has raised serious concerns about its infectiousness and antibody neutralization. However, few studies pay attention to the effect of the mutations on the dynamic interaction network of Omicron S protein trimers binding to the host angiotensin-converting enzyme 2 (ACE2). In this study, we conducted molecular dynamics (MD) simulations and enzyme linked immunosorbent assay (ELISA) to explore the binding strength and mechanism of wild type (WT), Delta, and Omicron S protein trimers to ACE2. The results showed that the binding capacities of both the two variants' S protein trimers to ACE2 are enhanced in varying degrees, indicating possibly higher cell infectiousness. Energy decomposition and protein-protein interaction network analysis suggested that both the mutational and conserved sites make effects on the increase in the overall affinity through a variety of interactions. The experimentally determined KD values by biolayer interferometry (BLI) and the predicted binding free energies of the RBDs of Delta and Omicron to mAb HLX70 revealed that the two variants may have the high risk of immune evasion from the mAb. These results are not only helpful in understanding the binding strength and mechanism of S protein trimer-ACE2 but also beneficial for drug, especially for antibody development.