Artificial intelligence–enabled virtual screening of ultra-large chemical libraries with deep docking

Nature Protocols, Год журнала: 2022, Номер 17(3), С. 672 - 697

Опубликована: Фев. 4, 2022

Язык: Английский

---

Molecular modeling in drug discovery

Informatics in Medicine Unlocked, Год журнала: 2022, Номер 29, С. 100880 - 100880

Опубликована: Янв. 1, 2022

With the financial requirements and high time associated with bringing a commercial drug to market, application of computer-aided design has been recognized as powerful technology in discovery pipeline. In accelerating discovery, molecular modeling techniques have experienced considerable growth computational capabilities over last decade. Pharmaceutical companies academic research organizations are currently using various lower cost required for an effective drug. this article, we focus on reviewing three key components (Molecular Docking, Molecular Dynamics, ADMET modeling), their applications, limitations small-molecule discovery. We discussed technicalities encircling dynamics docking, algorithms used develop docking softwares, models explored by these coupled scoring functions. also reviewed influence simulations (all atoms coarse-grained simulations) elucidated how ensembles generated from MD could pave way novel Furthermore, briefly explain role played pharmacokinetics pharmacodynamics profiling discovering new leads therapeutic efficacy. Besides success highlighted experimental corroboration silico discovered candidates. However, there is hardly market primarily use modeling, concluded review proposing possible solutions that foster advancement clinical drugs.

Язык: Английский

---

A critical overview of computational approaches employed for COVID-19 drug discovery

Chemical Society Reviews, Год журнала: 2021, Номер 50(16), С. 9121 - 9151

Опубликована: Янв. 1, 2021

COVID-19 has resulted in huge numbers of infections and deaths worldwide brought the most severe disruptions to societies economies since Great Depression. Massive experimental computational research effort understand characterize disease rapidly develop diagnostics, vaccines, drugs emerged response this devastating pandemic more than 130 000 COVID-19-related papers have been published peer-reviewed journals or deposited preprint servers. Much focused on discovery novel drug candidates repurposing existing against COVID-19, many such projects either exclusively computer-aided studies. Herein, we provide an expert overview key methods their applications for small-molecule therapeutics that reported literature. We further outline that, after first year pandemic, it appears not produced rapid global solutions. However, several known used clinic cure patients, a few repurposed continue be considered clinical trials, along with candidates. posit truly impactful tools must deliver actionable, experimentally testable hypotheses enabling combinations, open science sharing results are critical accelerate development novel, much needed COVID-19.

Язык: Английский

---

The transformational role of GPU computing and deep learning in drug discovery

Nature Machine Intelligence, Год журнала: 2022, Номер 4(3), С. 211 - 221

Опубликована: Март 23, 2022

Язык: Английский

---

Toward the Prediction of Binding Events in Very Flexible, Allosteric, Multidomain Proteins

Journal of Chemical Information and Modeling, Год журнала: 2025, Номер unknown

Опубликована: Фев. 5, 2025

Knowledge of the structures formed by proteins and small molecules is key to understand molecular principles chemotherapy for designing new more effective drugs. During early stage a drug discovery program, it customary predict ligand-protein complexes in silico, particularly when screening large compound databases. While virtual based on docking widely used this purpose, generally fails mimicking binding events associated with conformational changes protein, latter involve multiple domains. In work, we describe methodology generate bound-like conformations very flexible allosteric bearing sites exploiting only information unbound structure putative sites. The protocol validated paradigm enzyme adenylate kinase, which generated significant fraction structures. A these conformations, employed ensemble-docking calculations, allowed find native-like poses substrates inhibitors (binding active form enzyme), as well catalytically incompetent analogs inactive form). Our provides general framework generation challenging targets that are suitable host different ligands, demonstrating high sensitivity fine chemical details regulate protein's activity. We foresee applications screening, prediction impact amino acid mutations dynamics, protein engineering.

Язык: Английский

---

Defining and Exploring Chemical Spaces

Trends in Chemistry, Год журнала: 2020, Номер 3(2), С. 133 - 145

Опубликована: Дек. 17, 2020

Язык: Английский

---

Accelerating high-throughput virtual screening through molecular pool-based active learning

Chemical Science, Год журнала: 2021, Номер 12(22), С. 7866 - 7881

Опубликована: Янв. 1, 2021

Structure-based virtual screening is an important tool in early stage drug discovery that scores the interactions between a target protein and candidate ligands. As libraries continue to grow (in excess of 108 molecules), so too do resources necessary conduct exhaustive campaigns on these libraries. However, Bayesian optimization techniques, previously employed other scientific problems, can aid their exploration: surrogate structure-property relationship model trained predicted affinities subset library be applied remaining members, allowing least promising compounds excluded from evaluation. In this study, we explore application techniques computational docking datasets assess impact architecture, acquisition function, batch size performance. We observe significant reductions costs; for example, using directed-message passing neural network identify 94.8% or 89.3% top-50 000 ligands 100M member after testing only 2.4% upper confidence bound greedy strategy, respectively. Such model-guided searches mitigate increasing costs increasingly large accelerate high-throughput with applications beyond docking.

Язык: Английский

---

Biomolecular modeling thrives in the age of technology

Nature Computational Science, Год журнала: 2021, Номер 1(5), С. 321 - 331

Опубликована: Май 20, 2021

The biomolecular modeling field has flourished since its early days in the 1970s due to rapid adaptation and tailoring of state-of-the-art technology. resulting dramatic increase size timespan simulations outpaced Moore's law. Here, we discuss role knowledge-based versus physics-based methods hardware software advances propelling forward. This outreach suggests a bright future for modeling, where theory, experimentation simulation define three pillars needed address scientific biomedical challenges.

Язык: Английский

---

High-Throughput Virtual Screening and Validation of a SARS-CoV-2 Main Protease Noncovalent Inhibitor

Journal of Chemical Information and Modeling, Год журнала: 2021, Номер 62(1), С. 116 - 128

Опубликована: Ноя. 18, 2021

Despite the recent availability of vaccines against acute respiratory syndrome coronavirus 2 (SARS-CoV-2), search for inhibitory therapeutic agents has assumed importance especially in context emerging new viral variants. In this paper, we describe discovery a novel noncovalent small-molecule inhibitor, MCULE-5948770040, that binds to and inhibits SARS-Cov-2 main protease (Mpro) by employing scalable high-throughput virtual screening (HTVS) framework targeted compound library over 6.5 million molecules could be readily ordered purchased. Our HTVS leverages U.S. supercomputing infrastructure achieving nearly 91% resource utilization 126 docking calculations per hour. Downstream biochemical assays validate Mpro inhibitor with an inhibition constant (Ki) 2.9 μM (95% CI 2.2, 4.0). Furthermore, using room-temperature X-ray crystallography, show MCULE-5948770040 cleft primary binding site forming stable hydrogen bond hydrophobic interactions. We then used multiple μs-time scale molecular dynamics (MD) simulations machine learning (ML) techniques elucidate how bound ligand alters conformational states accessed Mpro, involving motions both proximal distal site. Together, our results demonstrate offers springboard further design.

Язык: Английский

---

Artificial Intelligence in Surveillance, Diagnosis, Drug Discovery and Vaccine Development against COVID-19

Pathogens, Год журнала: 2021, Номер 10(8), С. 1048 - 1048

Опубликована: Авг. 18, 2021

As of August 6th, 2021, the World Health Organization has notified 200.8 million laboratory-confirmed infections and 4.26 deaths from COVID-19, making it worst pandemic since 1918 flu. The main challenges in mitigating COVID-19 are effective vaccination, treatment, agile containment strategies. In this review, we focus on potential Artificial Intelligence (AI) surveillance, diagnosis, outcome prediction, drug discovery vaccine development. With help big data, AI tries to mimic cognitive capabilities a human brain, such as problem-solving learning abilities. Machine Learning (ML), subset AI, holds special promise for solving problems based experiences gained curated data. Advances methods have created an unprecedented opportunity building surveillance systems using deluge real-time data generated within short span time. During pandemic, many reports discussed utility approaches prioritization, delivery, supply chain drugs, vaccines, non-pharmaceutical interventions. This review will discuss clinical AI-based models also limitations faced by systems, model generalizability, explainability, trust pillars real-life deployment healthcare.

Язык: Английский

---