{"title":"SARS-CoV-2主蛋白酶QM/MM计算的可重复性","authors":"Xiaoli Sun,Ulf Ryde","doi":"10.1021/acs.jctc.5c00841","DOIUrl":null,"url":null,"abstract":"Combined quantum mechanics and molecular mechanics (QM/MM) calculations are a popular approach to study reaction mechanisms of enzymes. However, recently, the reproducibility of such calculations has been questioned, comparing the results of two software: NWChem and Q-Chem. Here, we continue and extend this study by including three additional software─ComQum, ORCA, and AMBER─using the same test case, the covalent attachment of the carmofur inhibitor to the catalytic Cys-145 residue of the SARS-CoV-2 main protease, using a quantum region of 83 atoms. We confirm that the various software programs give varying results for the reaction (ΔE) and activation (ΔE‡) energies. The main reason for the variation is how charges around the cleaved bonds between the QM and MM regions are treated, i.e., the charge-redistribution scheme. However, there are still differences of ∼10 kJ/mol between different implementations of the same method in ComQum and ORCA. Some of these problems can be solved by calculating the final energies with larger QM systems. We show that energies calculated with the big-QM approach are reasonably converged if atoms within 8 Å of the minimal QM region are included (∼1400 atoms), solvent-exposed charged residues are neutralized, and the calculation is performed in a continuum solvent with a dielectric constant of 80. On the other hand, we show that different setups of the protein lead to even larger differences in the calculated energies, by up to 114 kJ/mol. Even if the same approach is used and the only difference is how water molecules are added (by random) to the crystal structure, energies differ by 18-57 kJ/mol. The results also strongly depend on how much of the surrounding protein and solvent are relaxed in the calculations. Therefore, it seems that for a solvent-exposed active site, QM/MM calculations with minimized structures cannot be recommended. Instead, methods that incorporate dynamic effects and calculate free energies seem preferable.","PeriodicalId":45,"journal":{"name":"Journal of Chemical Theory and Computation","volume":"189 1","pages":""},"PeriodicalIF":5.7000,"publicationDate":"2025-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Reproducibility of QM/MM Calculations for the SARS-CoV-2 Main Protease.\",\"authors\":\"Xiaoli Sun,Ulf Ryde\",\"doi\":\"10.1021/acs.jctc.5c00841\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Combined quantum mechanics and molecular mechanics (QM/MM) calculations are a popular approach to study reaction mechanisms of enzymes. However, recently, the reproducibility of such calculations has been questioned, comparing the results of two software: NWChem and Q-Chem. Here, we continue and extend this study by including three additional software─ComQum, ORCA, and AMBER─using the same test case, the covalent attachment of the carmofur inhibitor to the catalytic Cys-145 residue of the SARS-CoV-2 main protease, using a quantum region of 83 atoms. We confirm that the various software programs give varying results for the reaction (ΔE) and activation (ΔE‡) energies. The main reason for the variation is how charges around the cleaved bonds between the QM and MM regions are treated, i.e., the charge-redistribution scheme. However, there are still differences of ∼10 kJ/mol between different implementations of the same method in ComQum and ORCA. Some of these problems can be solved by calculating the final energies with larger QM systems. We show that energies calculated with the big-QM approach are reasonably converged if atoms within 8 Å of the minimal QM region are included (∼1400 atoms), solvent-exposed charged residues are neutralized, and the calculation is performed in a continuum solvent with a dielectric constant of 80. On the other hand, we show that different setups of the protein lead to even larger differences in the calculated energies, by up to 114 kJ/mol. Even if the same approach is used and the only difference is how water molecules are added (by random) to the crystal structure, energies differ by 18-57 kJ/mol. The results also strongly depend on how much of the surrounding protein and solvent are relaxed in the calculations. Therefore, it seems that for a solvent-exposed active site, QM/MM calculations with minimized structures cannot be recommended. Instead, methods that incorporate dynamic effects and calculate free energies seem preferable.\",\"PeriodicalId\":45,\"journal\":{\"name\":\"Journal of Chemical Theory and Computation\",\"volume\":\"189 1\",\"pages\":\"\"},\"PeriodicalIF\":5.7000,\"publicationDate\":\"2025-07-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Chemical Theory and Computation\",\"FirstCategoryId\":\"92\",\"ListUrlMain\":\"https://doi.org/10.1021/acs.jctc.5c00841\",\"RegionNum\":1,\"RegionCategory\":\"化学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"CHEMISTRY, PHYSICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Chemical Theory and Computation","FirstCategoryId":"92","ListUrlMain":"https://doi.org/10.1021/acs.jctc.5c00841","RegionNum":1,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CHEMISTRY, PHYSICAL","Score":null,"Total":0}
Reproducibility of QM/MM Calculations for the SARS-CoV-2 Main Protease.
Combined quantum mechanics and molecular mechanics (QM/MM) calculations are a popular approach to study reaction mechanisms of enzymes. However, recently, the reproducibility of such calculations has been questioned, comparing the results of two software: NWChem and Q-Chem. Here, we continue and extend this study by including three additional software─ComQum, ORCA, and AMBER─using the same test case, the covalent attachment of the carmofur inhibitor to the catalytic Cys-145 residue of the SARS-CoV-2 main protease, using a quantum region of 83 atoms. We confirm that the various software programs give varying results for the reaction (ΔE) and activation (ΔE‡) energies. The main reason for the variation is how charges around the cleaved bonds between the QM and MM regions are treated, i.e., the charge-redistribution scheme. However, there are still differences of ∼10 kJ/mol between different implementations of the same method in ComQum and ORCA. Some of these problems can be solved by calculating the final energies with larger QM systems. We show that energies calculated with the big-QM approach are reasonably converged if atoms within 8 Å of the minimal QM region are included (∼1400 atoms), solvent-exposed charged residues are neutralized, and the calculation is performed in a continuum solvent with a dielectric constant of 80. On the other hand, we show that different setups of the protein lead to even larger differences in the calculated energies, by up to 114 kJ/mol. Even if the same approach is used and the only difference is how water molecules are added (by random) to the crystal structure, energies differ by 18-57 kJ/mol. The results also strongly depend on how much of the surrounding protein and solvent are relaxed in the calculations. Therefore, it seems that for a solvent-exposed active site, QM/MM calculations with minimized structures cannot be recommended. Instead, methods that incorporate dynamic effects and calculate free energies seem preferable.
期刊介绍:
The Journal of Chemical Theory and Computation invites new and original contributions with the understanding that, if accepted, they will not be published elsewhere. Papers reporting new theories, methodology, and/or important applications in quantum electronic structure, molecular dynamics, and statistical mechanics are appropriate for submission to this Journal. Specific topics include advances in or applications of ab initio quantum mechanics, density functional theory, design and properties of new materials, surface science, Monte Carlo simulations, solvation models, QM/MM calculations, biomolecular structure prediction, and molecular dynamics in the broadest sense including gas-phase dynamics, ab initio dynamics, biomolecular dynamics, and protein folding. The Journal does not consider papers that are straightforward applications of known methods including DFT and molecular dynamics. The Journal favors submissions that include advances in theory or methodology with applications to compelling problems.