Machine learning in preclinical drug discovery

Nature Chemical Biology, Journal Year: 2024, Volume and Issue: 20(8), P. 960 - 973

Published: July 19, 2024

Language: Английский

---

Prospective de novo drug design with deep interactome learning

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: April 22, 2024

Abstract De novo drug design aims to generate molecules from scratch that possess specific chemical and pharmacological properties. We present a computational approach utilizing interactome-based deep learning for ligand- structure-based generation of drug-like molecules. This method capitalizes on the unique strengths both graph neural networks language models, offering an alternative need application-specific reinforcement, transfer, or few-shot learning. It enables “zero-shot" construction compound libraries tailored bioactivity, synthesizability, structural novelty. In order proactively evaluate interactome framework protein design, potential new ligands targeting binding site human peroxisome proliferator-activated receptor (PPAR) subtype gamma are generated. The top-ranking designs chemically synthesized computationally, biophysically, biochemically characterized. Potent PPAR partial agonists identified, demonstrating favorable activity desired selectivity profiles nuclear receptors off-target interactions. Crystal structure determination ligand-receptor complex confirms anticipated mode. successful outcome positively advocates de application in bioorganic medicinal chemistry, enabling creation innovative bioactive

Language: Английский

---

Navigating the complexities of drug development for inflammatory bowel disease

Nature Reviews Drug Discovery, Journal Year: 2024, Volume and Issue: 23(7), P. 546 - 562

Published: May 22, 2024

Language: Английский

---

Integrating artificial intelligence in drug discovery and early drug development: a transformative approach

Biomarker Research, Journal Year: 2025, Volume and Issue: 13(1)

Published: March 14, 2025

Abstract Artificial intelligence (AI) can transform drug discovery and early development by addressing inefficiencies in traditional methods, which often face high costs, long timelines, low success rates. In this review we provide an overview of how to integrate AI the current process, as it enhance activities like target identification, discovery, clinical development. Through multiomics data analysis network-based approaches, help identify novel oncogenic vulnerabilities key therapeutic targets. models, such AlphaFold, predict protein structures with accuracy, aiding druggability assessments structure-based design. also facilitates virtual screening de novo design, creating optimized molecular for specific biological properties. development, supports patient recruitment analyzing electronic health records improves trial design through predictive modeling, protocol optimization, adaptive strategies. Innovations synthetic control arms digital twins reduce logistical ethical challenges simulating outcomes using real-world or data. Despite these advancements, limitations remain. models may be biased if trained on unrepresentative datasets, reliance historical lead overfitting lack generalizability. Ethical regulatory issues, privacy, challenge implementation AI. conclusion, a comprehensive about into processes. These efforts, although they will demand collaboration between professionals, robust quality, have transformative potential accelerate

Language: Английский

---

Flavonoids and Flavonoid-Based Nanopharmaceuticals as Promising Therapeutic Strategies for Colorectal Cancer—An Updated Literature Review

Pharmaceuticals, Journal Year: 2025, Volume and Issue: 18(2), P. 231 - 231

Published: Feb. 8, 2025

Colorectal cancer (CRC) represents one of the most serious health issues and third commonly diagnosed worldwide. However, treatment options for CRC are associated with adverse reactions, in some cases, resistance can develop. Flavonoids have emerged as promising alternatives prevention therapy due to their multitude biological properties ability target distinct processes involved pathogenesis. Their innate disadvantageous (e.g., low solubility stability, reduced bioavailability, lack tumor specificity) delayed potential inclusion flavonoids regimens but hastened design nanopharmaceuticals comprising a flavonoid agent entrapped nanosized delivery platform that not only counteract these inconveniences also provide an augmented therapeutic effect elevated safety profile by conferring targeted action. Starting brief presentation pathological features overview classes, present study comprehensively reviews anti-CRC activity different from mechanistic perspective while portraying latest discoveries made area flavonoid-containing nanocarriers proved efficient management. This review concludes showcasing future perspectives advancement flavonoid-based research.

Language: Английский

---

Harnessing pre-trained models for accurate prediction of protein-ligand binding affinity

BMC Bioinformatics, Journal Year: 2025, Volume and Issue: 26(1)

Published: Feb. 17, 2025

Abstract Background The binding between proteins and ligands plays a crucial role in the field of drug discovery. However, this area currently faces numerous challenges. On one hand, existing methods are constrained by limited availability labeled data, often performing inadequately when addressing complex protein-ligand interactions. other many models struggle to effectively capture flexible variations relative spatial relationships ligands. These issues not only significantly hinder advancement research but also adversely affect accuracy efficiency Therefore, response these challenges, our study aims enhance predictive capabilities through innovative approaches, providing more reliable support for discovery efforts. Methods This leverages pre-trained model with awareness prediction affinity. By perturbing structures small molecules manner consistent physical constraints employing self-supervised tasks, we improve representation molecule structures, allowing better adaptation affinity predictions. Meanwhile, approach enables identification potential sites on proteins. Results Our demonstrates higher correlation coefficient Extensive evaluation PDBBind v2019 refined set, CASF, Merck FEP benchmarks confirms model’s robustness strong generalization across diverse datasets. Additionally, achieves over 95% classification ROC site identification, underscoring its high pinpointing interaction regions. Conclusion presents novel that enhances predictions facilitates sites, showcasing computational design. Data code available at https://github.com/MIALAB-RUC/SableBind .

Language: Английский

---

Unveiling the antiviral inhibitory activity of ebselen and ebsulfur derivatives on SARS-CoV-2 using machine learning-based QSAR, LB-PaCS-MD, and experimental assay

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 26, 2025

Ebsulfur and ebselen derivatives that were proven to be potent inhibitors against the main protease (MPro) of SARS-CoV-2 which is an essential enzyme for viral replication chosen study quantitative structure–activity relationship (QSAR) analysis using a classical multiple linear regression (MLR) machine learning approach random forest (RF) artificial neural network (ANN) in order find between molecular structural properties biological inhibitory activities. With statistical criteria, R2 values MLR, RF, ANN models training set 0.83, 0.82, 0.92, respectively. The RMSE test considered model evaluation, results 0.27, 0.18, 0.09 models, Therefore, was best-obtained predicting MPro activity thirteen new synthetic analogs haven't tested assay before. Notably, our predicted activities then examined enzyme-based assays cytotoxicity tests, found compound P8 resulted good potential candidate activity. Furthermore, dynamics simulations performed dynamic interaction ligand binding site; showed pathway mechanism with key residues surrounding active site MPro, useful further development derivatives.

Language: Английский

---

Nanomedicines Targeting Metabolic Pathways in the Tumor Microenvironment: Future Perspectives and the Role of AI

Metabolites, Journal Year: 2025, Volume and Issue: 15(3), P. 201 - 201

Published: March 13, 2025

Background: Tumor cells engage in continuous self-replication by utilizing a large number of resources and capabilities, typically within an aberrant metabolic regulatory network to meet their own demands. This dysregulation leads the formation tumor microenvironment (TME) most solid tumors. Nanomedicines, due unique physicochemical properties, can achieve passive targeting certain tumors through enhanced permeability retention (EPR) effect, or active deliberate design optimization, resulting accumulation TME. The use nanomedicines target critical pathways holds significant promise. However, requires careful selection relevant drugs materials, taking into account multiple factors. traditional trial-and-error process is relatively inefficient. Artificial intelligence (AI) integrate big data evaluate delivery efficiency nanomedicines, thereby assisting nanodrugs. Methods: We have conducted detailed review key papers from databases, such as ScienceDirect, Scopus, Wiley, Web Science, PubMed, focusing on reprogramming, mechanisms action development metabolism, application AI empowering nanomedicines. integrated content present current status research metabolism potential future directions this field. Results: Nanomedicines possess excellent TME which be utilized disrupt cells, including glycolysis, lipid amino acid nucleotide metabolism. disruption selective killing disturbance Extensive has demonstrated that AI-driven methodologies revolutionized nanomedicine development, while concurrently enabling precise identification molecular regulators involved oncogenic reprogramming pathways, catalyzing transformative innovations targeted cancer therapeutics. Conclusions: great Additionally, will accelerate discovery metabolism-related targets, empower optimization help minimize toxicity, providing new paradigm for development.

Language: Английский

---

Comparative analysis of scaled entropies and topological properties of triphenylene-based metal and covalent organic frameworks

Chemical Papers, Journal Year: 2024, Volume and Issue: 78(7), P. 4095 - 4118

Published: April 7, 2024

Language: Английский

---

DrugFlow: An AI-Driven One-Stop Platform for Innovative Drug Discovery

Journal of Chemical Information and Modeling, Journal Year: 2024, Volume and Issue: 64(14), P. 5381 - 5391

Published: June 26, 2024

Artificial intelligence (AI)-aided drug design has demonstrated unprecedented effects on modern discovery, but there is still an urgent need for user-friendly interfaces that bridge the gap between these sophisticated tools and scientists, particularly those who are less computer savvy. Herein, we present DrugFlow, AI-driven one-stop platform offers a clean, convenient, cloud-based interface to streamline early discovery workflows. By seamlessly integrating range of innovative AI algorithms, covering molecular docking, quantitative structure-activity relationship modeling, generation, ADMET (absorption, distribution, metabolism, excretion toxicity) prediction, virtual screening, DrugFlow can offer effective solutions almost all crucial stages in including hit identification hit/lead optimization. We hope provide sufficiently valuable guidance aid real-word discovery. The available at https://drugflow.com.

Language: Английский

---