Eliminating the Deadwood: A Machine Learning Model for CCS Knowledge-Based Conformational Focusing for Lipids DOI Creative Commons
Mithony Keng, Kenneth M. Merz

Journal of Chemical Information and Modeling, Journal Year: 2024, Volume and Issue: 64(20), P. 7864 - 7872

Published: Oct. 8, 2024

Accurate elucidation of gas-phase chemical structures using collision cross section (CCS) values obtained from ion-mobility mass spectrometry benefits a synergism between experimental and in silico results. We have shown recent work that for molecule modest size with proscribed conformational space we can successfully capture conformation(s) match CCS values. However, flexible systems such as fatty acids many rotatable bonds multiple intramolecular London dispersion interactions, it becomes necessary to sample much greater space. Sampling more conformers, however, accrues significant computational cost downstream optimization steps involving quantum mechanics. To reduce this expense lipids, developed novel machine learning (ML) model facilitate conformer filtering according the estimated Herein report implementation our knowledge-based approach sampling resulted improved structure prediction agreement experiment by achieving favorable average errors ∼2% lipid both validation set test set. Moreover, most candidate conformations focusing achieved lower energy-minimum geometries than without focusing. Altogether, ML into modeling workflow has proven be beneficial quality results turnaround time. Finally, while is limited readily extended other molecules interest.

Language: Английский

Quantum mechanical-based strategies in drug discovery: Finding the pace to new challenges in drug design DOI Creative Commons
Tiziana Ginex, Javier Vázquez,

Carolina Estarellas

et al.

Current Opinion in Structural Biology, Journal Year: 2024, Volume and Issue: 87, P. 102870 - 102870

Published: June 24, 2024

The expansion of the chemical space to tangible libraries containing billions synthesizable molecules opens exciting opportunities for drug discovery, but also challenges power computer-aided design prioritize best candidates. This directly hits quantum mechanics (QM) methods, which provide chemically accurate properties, subject small-sized systems. Preserving accuracy while optimizing computational cost is at heart many efforts develop high-quality, efficient QM-based strategies, reflected in refined algorithms and approaches. QM-tailored physics-based force fields coupling QM with machine learning, conjunction computing performance supercomputing resources, will enhance ability use these methods discovery. challenge formidable, we undoubtedly see impressive advances that define a new era.

Language: Английский

Citations

11
Accurate Enthalpies of Formation for Bioactive Compounds from High-Level Ab Initio Calculations with Detailed Conformational Treatment: A Case of Cannabinoids DOI

Andrei F. Kazakov,

Eugene Paulechka

Journal of Chemical Theory and Computation, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 9, 2025

Our recently developed approach based on the local coupled-cluster with single, double, and perturbative triple excitation [LCCSD(T)] model gives very efficient means to compute ideal-gas enthalpies of formation. The expanded uncertainty (95% confidence) method is about 3 kJ·mol–1 for medium-sized compounds, comparable typical experimental measurements. Larger compounds interest often exhibit many conformations that can significantly differ in intramolecular interactions. Although present capabilities allow processing even a few hundred distinct conformer structures given compound, systems numbers well excess 1000. In this study, we investigate how reduce number expensive LCCSD(T) calculations large ensembles while controlling error approximation. best strategy found was correct results lower-level, surrogate (density functional theory, DFT) systematic manner. It also conformational contribution introduced by mainly driven (bias) rather than random component DFT energy deviation from target. This distinction usually overlooked benchmarking studies. As result work, formation 20 cannabinoid cannabinoid-related were obtained. Comprehensive analysis suggests uncertainties obtained values are below 4 kJ·mol–1.

Language: Английский

Citations

1
On the relevance of query definition in the performance of 3D ligand-based virtual screening DOI Creative Commons
Javier Vázquez, Ricardo López García,

Paula Llinares

et al.

Journal of Computer-Aided Molecular Design, Journal Year: 2024, Volume and Issue: 38(1)

Published: April 4, 2024

Abstract Ligand-based virtual screening (LBVS) methods are widely used to explore the vast chemical space in search of novel compounds resorting a variety properties encoded 1D, 2D or 3D descriptors. The success 3D-LBVS is affected by overlay molecular pairs, thus making selection template compound, accessible conformational and choice query conformation be potential factors that modulate successful retrieval actives. This study examines impact adopting different choices for template, paying also attention influence exerted structural similarity between templates analysis performed using PharmScreen, LBVS tool relies on measurements hydrophobic/philic pattern molecules, Phase Shape, which based alignment atom triplets followed refinement volume overlap. original DUD-E + database Morgan Fingerprint filtered version (denoted -Diverse; available https://github.com/Pharmacelera/Query-models-to-3DLBVS ), was prepared minimize resemblance Although most cases exhibits mild overall performance, critical made disclose factors, such as content features actives induction strain underlie drastic definition recovery certain targets. findings this research provide valuable guidance assisting campaigns. Graphical

Language: Английский

Citations

4
Partial to Total Generation of 3D Transition-Metal Complexes DOI Creative Commons
H.-Q. Jin, Kenneth M. Merz

Journal of Chemical Theory and Computation, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 9, 2024

The design of transition-metal complexes (TMCs) has drawn much attention over the years because their important applications as metallodrugs and functional materials. In this work, we present an extension our recently reported approach, LigandDiff [Jin et al.

Language: Английский

Citations

3
Synthon-Based Strategies Exploiting Molecular Similarity and Protein–Ligand Interactions for Efficient Screening of Ultra-Large Chemical Libraries DOI
Brian Medel-Lacruz, Albert Herrero,

Fernando A. Martín

et al.

Journal of Chemical Information and Modeling, Journal Year: 2025, Volume and Issue: unknown

Published: April 28, 2025

The rapid expansion of ultralarge chemical libraries has revolutionized drug discovery, providing access to billions compounds. However, this growth poses relevant challenges for traditional virtual screening (VS) methods. To address these limitations, synthon-based approaches have emerged as scalable alternatives, exploiting combinatorial chemistry principles prioritize building blocks over enumerated molecules. In work, we present exaScreen and exaDock, two novel methodologies designed ligand-based structure-based VS, respectively. the former case, synthon selection is guided by 3D hydrophobic/philic distribution pattern in conjunction with a specific alignment protocol based on quadrupolar atoms that participate linking bonds between fragments. On other hand, accommodation binding site under geometrically restrained docking hybrid compounds used optimal combinations. These strategies exhibit comparable performance search performed using fully identifying active significantly lower computational cost, offering computationally efficient VS spaces.

Language: Английский

Citations

0
Data-Driven Virtual Screening of Conformational Ensembles of Transition-Metal Complexes DOI Creative Commons

Sára Finta,

Adarsh V. Kalikadien, Evgeny A. Pidko

et al.

Journal of Chemical Theory and Computation, Journal Year: 2025, Volume and Issue: unknown

Published: May 9, 2025

Transition-metal complexes serve as highly enantioselective homogeneous catalysts for various transformations, making them valuable in the pharmaceutical industry. Data-driven prediction models can accelerate high-throughput catalyst design but require computer-readable representations that account conformational flexibility. This is typically achieved through high-level conformer searches, followed by DFT optimization of transition-metal complexes. However, selection remains reliant on human assumptions, with no cost-efficient and generalizable workflow available. To address this, we introduce an automated approach to correlate CREST(GFN2-xTB//GFN-FF)-generated ensembles their DFT-optimized counterparts systematic selection. We analyzed 24 precatalyst structures, performing CREST full optimization. Three filtering methods were evaluated: (i) geometric ligand descriptors, (ii) PCA-based selection, (iii) DBSCAN clustering using RMSD energy. The proposed validated Rh-based featuring bisphosphine ligands, which are widely employed hydrogenation reactions. assess general applicability, both its corresponding acrylate-bound complex analyzed. Our results confirm overestimates flexibility, energy-based ineffective. failed distinguish conformers energy, while RMSD-based improved lacked tunability. provided most effective approach, eliminating redundancies preserving key configurations. method remained robust across data sets computationally efficient without requiring molecular descriptor calculations. These findings highlight limitations advantages structure-based approaches While a practical solution, parameters remain system-dependent. For high-accuracy applications, refined energy calculations may be necessary; however, DBSCAN-based offers accessible strategy rapid involving

Language: Английский

Citations

0
Toward AI/ML-assisted discovery of transition metal complexes DOI
H.-Q. Jin, Kenneth M. Merz

Annual reports in computational chemistry, Journal Year: 2024, Volume and Issue: unknown

Published: Jan. 1, 2024

Language: Английский

Citations

1
Eliminating the Deadwood: A Machine Learning Model for CCS Knowledge-Based Conformational Focusing for Lipids DOI Creative Commons
Mithony Keng, Kenneth M. Merz

Journal of Chemical Information and Modeling, Journal Year: 2024, Volume and Issue: 64(20), P. 7864 - 7872

Published: Oct. 8, 2024

Accurate elucidation of gas-phase chemical structures using collision cross section (CCS) values obtained from ion-mobility mass spectrometry benefits a synergism between experimental and in silico results. We have shown recent work that for molecule modest size with proscribed conformational space we can successfully capture conformation(s) match CCS values. However, flexible systems such as fatty acids many rotatable bonds multiple intramolecular London dispersion interactions, it becomes necessary to sample much greater space. Sampling more conformers, however, accrues significant computational cost downstream optimization steps involving quantum mechanics. To reduce this expense lipids, developed novel machine learning (ML) model facilitate conformer filtering according the estimated Herein report implementation our knowledge-based approach sampling resulted improved structure prediction agreement experiment by achieving favorable average errors ∼2% lipid both validation set test set. Moreover, most candidate conformations focusing achieved lower energy-minimum geometries than without focusing. Altogether, ML into modeling workflow has proven be beneficial quality results turnaround time. Finally, while is limited readily extended other molecules interest.

Language: Английский

Citations

0