Advancing Drug Discovery via Artificial Intelligence

Trends in Pharmacological Sciences, Journal Year: 2019, Volume and Issue: 40(8), P. 592 - 604

Published: July 15, 2019

Language: Английский

---

Use of Zebrafish in Drug Discovery Toxicology

Chemical Research in Toxicology, Journal Year: 2019, Volume and Issue: 33(1), P. 95 - 118

Published: Oct. 18, 2019

Unpredicted human safety events in clinical trials for new drugs are costly terms of health and money. The drug discovery industry attempts to minimize those with diligent preclinical testing. Current standard practices good at preventing toxic compounds from being tested the clinic; however, false negative toxicity results still a reality. Continual improvement must be pursued realm. Higher-quality therapies can brought forward more information about potential toxicities associated mechanisms. zebrafish model is bridge between vitro assays mammalian vivo studies. This powerful its breadth application tractability research. In past two decades, our understanding disease biology has grown significantly owing thousands studies on this tiny vertebrate. Review summarizes challenges strengths model, discusses 3Rs value that it deliver, highlights translatable untranslatable biology, brings together reports recent focusing toxicology.

Language: Английский

---

Bioisosteres of the Phenyl Ring: Recent Strategic Applications in Lead Optimization and Drug Design

Journal of Medicinal Chemistry, Journal Year: 2021, Volume and Issue: 64(19), P. 14046 - 14128

Published: Sept. 30, 2021

The benzene moiety is the most prevalent ring system in marketed drugs, underscoring its historic popularity drug design either as a pharmacophore or scaffold that projects pharmacophoric elements. However, introspective analyses of medicinal chemistry practices at beginning 21st century highlighted indiscriminate deployment phenyl rings an important contributor to poor physicochemical properties advanced molecules, which limited their prospects being developed into effective drugs. This Perspective deliberates on and applications bioisosteric replacements for have provided practical solutions range developability problems frequently encountered lead optimization campaigns. While effect compound contextual nature, substitution can enhanced potency, solubility, metabolic stability while reducing lipophilicity, plasma protein binding, phospholipidosis potential, inhibition cytochrome P450 enzymes hERG channel.

Language: Английский

---

Combine and conquer: challenges for targeted therapy combinations in early phase trials

Nature Reviews Clinical Oncology, Journal Year: 2016, Volume and Issue: 14(1), P. 57 - 66

Published: July 5, 2016

Language: Английский

---

Computational analysis of calculated physicochemical and ADMET properties of protein-protein interaction inhibitors

Scientific Reports, Journal Year: 2017, Volume and Issue: 7(1)

Published: April 11, 2017

Abstract The modulation of PPIs by low molecular weight chemical compounds, particularly orally bioavailable molecules, would be very valuable in numerous disease indications. However, it is known that PPI inhibitors (iPPIs) tend to have properties are linked poor Absorption, Distribution, Metabolism, Excretion and Toxicity (ADMET) some cases clinical outcomes. Previously reported silico analyses iPPIs essentially focused on physicochemical but several other ADMET parameters important assess. In order gain new insights into the iPPIs, computations were carried out eight datasets collected from databases. These involve compounds targeting enzymes, GPCRs, ion channels, nuclear receptors, allosteric modulators, oral marketed drugs, natural product-derived drugs iPPIs. Several trends should assist design optimization future inhibitors, either for drug discovery endeavors or biology projects.

Language: Английский

---

Managing the challenge of drug-induced liver injury: a roadmap for the development and deployment of preclinical predictive models

Nature Reviews Drug Discovery, Journal Year: 2019, Volume and Issue: 19(2), P. 131 - 148

Published: Nov. 20, 2019

Language: Английский

---

LightGBM: An Effective and Scalable Algorithm for Prediction of Chemical Toxicity–Application to the Tox21 and Mutagenicity Data Sets

Journal of Chemical Information and Modeling, Journal Year: 2019, Volume and Issue: 59(10), P. 4150 - 4158

Published: Sept. 27, 2019

Machine learning algorithms have attained widespread use in assessing the potential toxicities of pharmaceuticals and industrial chemicals because their faster speed lower cost compared to experimental bioassays. Gradient boosting is an effective algorithm that often achieves high predictivity, but historically relative long computational time limited its applications predicting large compound libraries or developing silico predictive models require frequent retraining. LightGBM, a recent improvement gradient algorithm, inherited predictivity resolved scalability by adopting leaf-wise tree growth strategy introducing novel techniques. In this study, we performance LightGBM deep neural networks, random forests, support vector machines, XGBoost. All were rigorously evaluated on publicly available Tox21 mutagenicity data sets using Bayesian optimization integrated nested 10-fold cross-validation scheme performs hyperparameter while examining model generalizability transferability new data. The evaluation results demonstrated highly scalable offering best consuming significantly shorter than other investigated across all sets. We recommend for safety assessment also areas cheminformatics fulfill ever-growing demand accurate rapid prediction various toxicity activity related end points present pharmaceutical chemical industry.

Language: Английский

---

An Overview of Machine Learning and Big Data for Drug Toxicity Evaluation

Chemical Research in Toxicology, Journal Year: 2019, Volume and Issue: 33(1), P. 20 - 37

Published: Oct. 18, 2019

Drug toxicity evaluation is an essential process of drug development as it reportedly responsible for the attrition approximately 30% candidates. The rapid increase in number and types large toxicology data sets together with advances computational methods may be used to improve many steps safety evaluation. silico models screen understand mechanisms particularly beneficial early stages where assessment can most reduce expenses labor time. To facilitate this, machine learning have been employed evaluate but are often limited by small less diverse sets. Recent big such molecular descriptors, toxicogenomics, high-throughput bioactivity help alleviate some current challenges. In this article, common reviewed examples studies that methodology. Furthermore, a comprehensive overview different tools available build prediction has provided give landscape highlight opportunities challenges related them.

Language: Английский

---

Improving target assessment in biomedical research: the GOT-IT recommendations

Nature Reviews Drug Discovery, Journal Year: 2020, Volume and Issue: 20(1), P. 64 - 81

Published: Nov. 16, 2020

Language: Английский

---

Molecular inflation, attrition and the rule of five

Advanced Drug Delivery Reviews, Journal Year: 2016, Volume and Issue: 101, P. 22 - 33

Published: Feb. 1, 2016

Language: Английский

---