Special Features of COVID-19 in the FMODB: Fragment Molecular Orbital Calculations and Interaction Energy Analysis of SARS-CoV-2-Related Proteins

Fault-tolerant quantum chemical calculations with improved machine-learning models

Easy and effective usage of computational resources is crucial for scientific calculations. Following our recent work of machine-learning (ML) assisted scheduling optimization [J. Comput. Chem. 2023, 44, 1174], we further propose (1) the improved ML models for the better predictions of computational loads, and as such, more elaborate load-balancing calculations can be expected; (2) the idea of coded computation, that is, the integration of gradient coding, in order to introduce fault tolerance during the distributed calculations; and (3) their applications together with re-normalized exciton model with time-dependent density functional theory (REM-TDDFT) for calculating the excited states. Illustrated benchmark calculations include P38 protein, and solvent model with one or several excitable centers. The results show that the improved ML-assisted coded calculations can further improve the load-balancing and cluster utilization, owing primarily profit in fault tolerance that aims at the automated quantum chemical calculations for both ground and excited states.

Quantum chemical calculation dataset for representative protein folds by the fragment molecular orbital method

The function of a biomacromolecule is not only determined by its three-dimensional structure but also by its electronic state. Quantum chemical calculations are promising non-empirical methods available for determining the electronic state of a given structure. In this study, we used the fragment molecular orbital (FMO) method, which applies to biopolymers such as proteins, to provide physicochemical property values on representative structures in the SCOP2 database of protein families, a subset of the Protein Data Bank. Our dataset was constructed by over 5,000 protein structures, including over 200 million inter-fragment interaction energies (IFIEs) and their energy components obtained by pair interaction energy decomposition analysis (PIEDA) using FMO-MP2/6-31 G*. Moreover, three basis sets, 6-31 G*, 6-31 G**, and cc-pVDZ, were used for the FMO calculations of each structure, making it possible to compare the energies obtained with different basis functions for the same fragment pair. The total data size is approximately 6.7 GB. Our dataset will be useful for functional analyses and machine learning based on the physicochemical property values of proteins.

Effect of Direct and Water-Mediated Interactions on the Identification of Hotspots in Biomolecular Complexes with Multiple Subsystems

Identification of important residues in biochemical complexes is often a crucial step for many problems in molecular biology and biochemistry. A method is proposed to identify hotspots in biomolecular complexes based on a new metric, derived from networks representing molecular subunits (residues, bridging solvent molecules, ligands etc.) connected by interactions. A singular value decomposition of the weighted adjacency matrix is used to construct a scalar rank for each subunit that reflects its importance in the residue interaction network. This metric is called the singular value centrality. In addition, a new formalism is proposed to account for water-mediated interactions in the ranking of residues. Interactions for a residue network can be provided by various computational methods. In this work interactions are obtained from full quantum-mechanical calculations of protein-protein complexes using the fragment molecular orbital method. The ranking results are shown to be in good agreement with earlier computational and experimental studies. The developed method can be used to gain a deeper insight into the role of subunits in complex biomolecular systems.

FMOe: Preprocessing and Visualizing Package of the Fragment Molecular Orbital Method for Molecular Operating Environment and Its Applications in Covalent Ligand and Metalloprotein Analyses

The fragment molecular orbital (FMO) method is an efficient quantum chemical calculation technique for large biomolecules, dividing each into smaller fragments and providing interfragment interaction energies (IFIEs) that support our understanding of molecular recognition. The ab initio fragment MO method (ABINIT-MP), an FMO processing program, can automatically divide typical proteins and nucleic acids. In contrast, small molecules such as ligands and heterosystems must be manually divided. Thus, we developed a graphical user interface to easily handle such manual fragmentation as a library for the Molecular Operating Environment (MOE) that preprocesses and visualizes FMO calculations. We demonstrated fragmentation with IFIE analyses for the two following cases: (1) covalent cysteine-ligand bonding inside the SARS-CoV-2 main protease (Mpro) and nirmatrelvir (Paxlovid) complex and (2) the metal coordination inside a zinc-bound cyclic peptide. IFIE analysis successfully identified the key amino acid residues for the molecular recognition of nirmatrelvir with Mpro and the details of their interactions (e.g., hydrogen bonds and CH/π interactions) via ligand fragmentation of functional group units. In metalloproteins, we found an efficient and accurate scheme for the fragmentation of Zn2+ ions with four histidines coordinated to the ion. FMOe simplifies manual fragmentation, allowing users to experiment with various fragmentation patterns and perform in-depth IFIE analysis with high accuracy. In the future, our findings will provide valuable insight into complicated cases, such as ligand fragmentation in modality drug discovery, especially for medium-sized molecules and metalloprotein fragmentation around metals.

Prediction of Binding Pose and Affinity of Nelfinavir, a SARS-CoV-2 Main Protease Repositioned Drug, by Combining Docking, Molecular Dynamics, and Fragment Molecular Orbital Calculations

A novel in silico drug design procedure is described targeting the Main protease (Mpro) of the SARS-CoV-2 virus. The procedure combines molecular docking, molecular dynamics (MD), and fragment molecular orbital (FMO) calculations. The binding structure and properties of Mpro were predicted for Nelfinavir (NFV), which had been identified as a candidate compound through drug repositioning, targeting Mpro. Several poses of the Mpro and NFV complexes were generated by docking, from which four docking poses were selected by scoring with FMO energy. Then, each pose was subjected to MD simulation, 100 snapshot structures were sampled from each of the generated MD trajectories, and the structures were evaluated by FMO calculations to rank the pose based on binding energy. Several residues were found to be important in ligand recognition, including Glu47, Asp48, Glu166, Asp187, and Gln189, all of which interacted strongly with NFV. Asn142 is presumably regarded to form hydrogen bonds or CH/π interaction with NFV; however, in the present calculation, their interactions were transient. Moreover, the tert-butyl group of NFV had no interaction with Mpro. Identifying such strong and weak interactions provides candidates for maintaining and substituting ligand functional groups and important suggestions for drug discovery using drug repositioning. Besides the interaction between NFV and the amino acid residues of Mpro, the desolvation effect of the binding pocket also affected the ranking order. A similar procedure of drug design was applied to Lopinavir, and the calculated interaction energy and experimental inhibitory activity value trends were consistent. Our approach provides a new guideline for structure-based drug design starting from a candidate compound whose complex crystal structure has not been obtained.

Machine-learning assisted scheduling optimization and its application in quantum chemical calculations

Easy and effective usage of computational resources is crucial for scientific calculations, both from the perspectives of timeliness and economic efficiency. This work proposes a bi-level optimization framework to optimize the computational sequences. Machine-learning (ML) assisted static load-balancing, and different dynamic load-balancing algorithms can be integrated. Consequently, the computational and scheduling engine of the ParaEngine is developed to invoke optimized quantum chemical (QC) calculations. Illustrated benchmark calculations include high-throughput drug suit, solvent model, P38 protein, and SARS-CoV-2 systems. The results show that the usage rate of given computational resources for high throughput and large-scale fragmentation QC calculations can primarily profit, and faster accomplishing computational tasks can be expected when employing high-performance computing (HPC) clusters.

Quantum Chemical Interaction Analysis between SARS-CoV-2 Main Protease and Ensitrelvir Compared with Its Initial Screening Hit

A non-covalent oral drug targeting SARS-CoV-2 main protease (Mpro), ensitrelvir (Xocova), has been developed using structure-based drug design (SBDD). To elucidate the factors responsible for enhanced inhibitory activities from an in silico screening hit compound to ensitrelvir, we analyzed the interaction energies of the inhibitors with each residue of Mpro using fragment molecular orbital (FMO) calculations. This analysis reveals that functional group conversion for P1' and P1 parts in the inhibitors increases the strength of existing interactions with Mpro and also provides novel interactions for ensitrelvir; the associated changes in the conformation of Mpro induce further interactions for ensitrelvir in other parts, including hydrogen bonds, a halogen bond, and π-orbital interactions. Thus, we illuminate the promising strategies of SBDD for leading ensitrelvir to get higher activity against Mpro by elucidating microscopic interactions through FMO-based analysis. These detailed mechanism findings, including water cross-linkings, will help to design novel inhibitors in SBDD.

System truncation accelerates binding affinity calculations with the fragment molecular orbital method: A benchmark study

The fragment molecular orbital (FMO) method is a fast quantum-mechanical method that divides systems into pieces of fragments and performs ab initio calculations. The system truncation enables further speed improvement. In this article, we systematically study the effects of system truncations on binding affinity calculations obtained with FMO in combination with either the polarizable continuum model (FMO/PCM) or in combination with the Møller-Plesset method (FMO-MP2). We have used five protein complexes with ligands of several charged states. The calculated binding energies of the size variants of the truncated system, including only a restricted number of atoms around the ligand, are compared to the energy obtained from a full system. The result shows that the systems could be truncated to a radius of 8 Å from neutral ligands within an error of 0.7 kcal/mol, and 12 Å from charged ligands within an error of 1.1 kcal/mol for calculating the binding energy in solution.

Hotspot Identification and Drug Design of Protein-Protein Interaction Modulators Using the Fragment Molecular Orbital Method

Protein-protein interactions (PPIs) are essential for the function of many proteins. Aberrant PPIs have the potential to lead to disease, making PPIs promising targets for drug discovery. There are over 64,000 PPIs in the human interactome reference database; however, to date, very few PPI modulators have been approved for clinical use. Further development of PPI-specific therapeutics is highly dependent on the availability of structural data and the existence of reliable computational tools to explore the interface between two interacting proteins. The fragment molecular orbital (FMO) quantum mechanics method offers comprehensive and computationally inexpensive means of identifying the strength (in kcal/mol) and the chemical nature (electrostatic or hydrophobic) of the molecular interactions taking place at the protein-protein interface. We have integrated FMO and PPI exploration (FMO-PPI) to identify the residues that are critical for protein-protein binding (hotspots). To validate this approach, we have applied FMO-PPI to a dataset of protein-protein complexes representing several different protein subfamilies and obtained FMO-PPI results that are in agreement with published mutagenesis data. We observed that critical PPIs can be divided into three major categories: interactions between residues of two proteins (intermolecular), interactions between residues within the same protein (intramolecular), and interactions between residues of two proteins that are mediated by water molecules (water bridges). We extended our findings by demonstrating how this information obtained by FMO-PPI can be utilized to support the structure-based drug design of PPI modulators (SBDD-PPI).

Fragment molecular orbital calculations for biomolecules

Exploring biomolecule behavior, such as proteins and nucleic acids, using quantum mechanical theory can identify many life science phenomena from first principles. Fragment molecular orbital (FMO) calculations of whole single particles of biomolecules can determine the electronic state of the interior and surface of molecules and explore molecular recognition mechanisms based on intermolecular and intramolecular interactions. In this review, we summarized the current state of FMO calculations in drug discovery, virology, and structural biology, as well as recent developments from data science.