Machine Learning for Toxicity Prediction Using Chemical Structures: Pillars for Success in the Real World

Chemical Research in Toxicology, Год журнала: 2025, Номер unknown

Опубликована: Май 2, 2025

Machine learning (ML) is increasingly valuable for predicting molecular properties and toxicity in drug discovery. However, toxicity-related end points have always been challenging to evaluate experimentally with respect vivo translation due the required resources human animal studies; this has impacted data availability field. ML can augment or even potentially replace traditional experimental processes depending on project phase specific goals of prediction. For instance, models be used select promising compounds on-target effects deselect those undesirable characteristics (e.g., off-target ineffective unfavorable pharmacokinetics). reliance not without risks, biases stemming from nonrepresentative training data, incompatible choice algorithm represent underlying poor model building validation approaches. This might lead inaccurate predictions, misinterpretation confidence ultimately suboptimal decision-making. Hence, understanding predictive validity utmost importance enable faster development timelines while improving quality decisions. perspective emphasizes need enhance application machine discovery, focusing well-defined sets prediction based small molecule structures. We focus five crucial pillars success ML-driven property prediction: (1) set selection, (2) structural representations, (3) algorithm, (4) validation, (5) predictions Understanding these key will foster collaboration coordination between researchers toxicologists, which help advance discovery development.

Язык: Английский

---

Ligand-Induced Biased Activation of GPCRs: Recent Advances and New Directions from In Silico Approaches

Molecules, Год журнала: 2025, Номер 30(5), С. 1047 - 1047

Опубликована: Фев. 25, 2025

G-protein coupled receptors (GPCRs) are the largest family of membrane proteins engaged in transducing signals from extracellular environment into cell. GPCR-biased signaling occurs when two different ligands, sharing same binding site, induce distinct pathways. This selective offers significant potential for design safer and more effective drugs. Although its molecular mechanism remains elusive, big efforts made to try explain this using a wide range methods. Recent advances computational techniques AI technology have introduced variety simulations machine learning tools that facilitate modeling GPCR signal transmission analysis ligand-induced biased signaling. In review, we present current state silico approaches elucidate structural includes dynamics capture main interactions causing bias. We also highlight major contributions impacts transmembrane domains, loops, mutations mediating Moreover, discuss impact models on bias prediction diffusion-based generative ligands. Ultimately, review addresses future directions studying problem through approaches.

Язык: Английский

---

DrugDiff: small molecule diffusion model with flexible guidance towards molecular properties

Journal of Cheminformatics, Год журнала: 2025, Номер 17(1)

Опубликована: Фев. 25, 2025

Abstract With the cost/yield-ratio of drug development becoming increasingly unfavourable, recent work has explored machine learning to accelerate early stages process. Given current success deep generative models across domains, we here investigated their application property-based proposal new small molecules for development. Specifically, trained a latent diffusion model— DrugDiff —paired with predictor guidance generate novel compounds variety desired molecular properties. The architecture was designed be highly flexible and easily adaptable future scenarios. Our experiments showed successful generation unique, diverse targeted code is available at https://github.com/MarieOestreich/DrugDiff . Scientific Contribution This expands use modelling in field from previously introduced proteins RNA presented molecules. making up majority drugs, but simultaneously being difficult model due elaborate chemical rules, this tackles level difficulty comparison sequence-based molecule as case RNA. Additionally, demonstrated framework flexible, allowing easy addition or removal considered properties without need retrain model, it research settings shows compelling performance wide

Язык: Английский

---

Large Model Era: Deep Learning in Osteoporosis Drug Discovery

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

Опубликована: Фев. 26, 2025

Osteoporosis is a systemic microstructural degradation of bone tissue, often accompanied by fractures, pain, and other complications, resulting in decline patients' life quality. In response to the increased incidence osteoporosis, related drug discovery has attracted more attention, but it faced with challenges due long development cycle high cost. Deep learning powerful data processing capabilities shown significant advantages field discovery. With technology, applied all stages particular, large models, which have been developed rapidly recently, provide new methods for understanding disease mechanisms promoting because their parameters ability deal complex tasks. This review introduces traditional models deep domain, systematically summarizes applications each stage discovery, analyzes application prospect osteoporosis Finally, limitations are discussed depth, order help future

Язык: Английский

---

How Generative Artificial Intelligence Can Transform Drug Discovery?

European Journal of Medicinal Chemistry, Год журнала: 2025, Номер 295, С. 117825 - 117825

Опубликована: Май 27, 2025

Язык: Английский

---

Machine Learning Transition State Geometries and Applications in Reaction Property Prediction

Wiley Interdisciplinary Reviews Computational Molecular Science, Год журнала: 2025, Номер 15(3)

Опубликована: Май 1, 2025

ABSTRACT The calculation of transition state (TS) geometries is essential for understanding reaction mechanisms and rational synthetic methodology design. However, traditional methods like density functional theory are often too computationally expensive large‐scale TS identification significantly slower than high‐throughput experimental screening methods. Recent advancements in machine learning (ML) offer promising alternatives, enabling the direct prediction geometries, reducing reliance on quantum mechanical (QM) calculations, affording predictions ahead experiments. works explored here include broader application ML property prediction, emphasizing how accurate can serve as vital input data to improve model accuracy. A comprehensive review developed explicitly predict then presented, with attention their downstream tasks, such energy barrier use initial structures further optimization via QM Finally, a critical evaluation accuracy limitations existing discussed, highlighting challenges that impede wider adoption areas where research needed.

Язык: Английский

---

Computer-aided Molecular Design by Aligning Generative Diffusion Models: Perspectives and Challenges

Computers & Chemical Engineering, Год журнала: 2024, Номер 194, С. 108989 - 108989

Опубликована: Дек. 27, 2024

Язык: Английский

---

Diffusion Generative Models for Designing Efficient Singlet Fission Dimers

The Journal of Physical Chemistry A, Год журнала: 2024, Номер 129(1), С. 407 - 414

Опубликована: Дек. 30, 2024

Diffusion generative models, a class of machine learning techniques, have shown remarkable promise in materials science and chemistry by enabling the precise generation complex molecular structures. In this article, we propose novel application diffusion models for stabilizing reactive structures identified through quantum mechanical screening. Specifically, focus on design challenge presented singlet fission (SF), phenomenon crucial advancing solar cell efficiency beyond theoretical limits. While has been successful predicting intermolecular arrangements with enhanced SF coupling, practical implementation these configurations faces challenges due to discrepancies between favorable stabilized To address gap, introduce three-step strategy combining screening identifying optimal linkers. Through case study cibalackrot dimers, promising material, demonstrate efficacy our approach enhancing desired arrangements.

Язык: Английский

---