RNA-targeted small-molecule drug discoveries: a machine-learning perspective

RNA Biology, Год журнала: 2023, Номер 20(1), С. 384 - 397

Опубликована: Июнь 19, 2023

In the past two decades, machine learning (ML) has been extensively adopted in protein-targeted small molecule (SM) discovery. Once trained, ML models could exert their predicting abilities on large volumes of molecules within a short time. However, applying approaches to discover RNA-targeted SMs is still its early stages. This primarily because intrinsic structural instability RNA that impede structure-based screening or designing SMs. Recently, with more studies revealing structures and growing number ligands being identified, it resulted an increased interest field drugging RNA. Undeniably, intracellular much abundant than protein and, if successfully targeted, will be major alternative target for therapeutics. Therefore, this context, as well under premise having RNA-related research data, ML-based methods can get involved improving speed traditional experimental processes. [Figure: see text].

Язык: Английский

---

The emergent role of explainable artificial intelligence in the materials sciences

Cell Reports Physical Science, Год журнала: 2023, Номер 4(10), С. 101630 - 101630

Опубликована: Окт. 1, 2023

The combination of rational machine learning with creative materials science makes informatics a powerful way discovering, designing, and screening new materials. However, moving from promising prediction to practical strategy often requires more than just an instructive structure-property relationship; understanding how method uses the structural feature predict target properties becomes critical. Explainable artificial intelligence (XAI) is emerging field in computer based statistics that can augment workflows. XAI be used as forensic analysis understand consequences data, model, application decisions or model refinement capable distinguishing important features nuisance variables. Here, we outline state art highlight methods most useful physical sciences. This guide focuses on characteristics are relevant will become increasingly researchers move toward using deeper neural networks large language models.

Язык: Английский

---

Engineered DNA bonsai system for ultrasensitive wide-field determination and intracellular dynamic imaging of protein with tunable dynamic range and sensitivity

Nano Today, Год журнала: 2024, Номер 54, С. 102111 - 102111

Опубликована: Янв. 4, 2024

Язык: Английский

---

Targeting RNA with small molecules, from RNA structures to precision medicines: IUPHAR review: 40

British Journal of Pharmacology, Год журнала: 2024, Номер 181(21), С. 4152 - 4173

Опубликована: Сен. 3, 2024

RNA plays important roles in regulating both health and disease biology all kingdoms of life. Notably, can form intricate three‐dimensional structures, their biological functions are dependent on these structures. Targeting the structured regions with small molecules has gained increasing attention over past decade, because it provides chemical probes to study fundamental processes lead medicines for diseases unmet medical needs. Recent advances structure prediction determination have accelerated rational design development RNA‐targeted modulate pathology. However, challenges remain advancing towards clinical applications. This review summarizes strategies identify recognizing augment functionality RNA‐binding molecules. We focus recent developing as potential therapeutics a variety diseases, encompassing different modes actions targeting strategies. Furthermore, we present current gaps between early‐stage discovery applications, well roadmap overcome near future.

Язык: Английский

---

G4-QuadScreen: A Computational Tool for Identifying Multi-Target-Directed Anticancer Leads against G-Quadruplex DNA

Cancers, Год журнала: 2023, Номер 15(15), С. 3817 - 3817

Опубликована: Июль 27, 2023

The study presents ‘G4-QuadScreen’, a user-friendly computational tool for identifying MTDLs against G4s. Also, it offers few hit based on in silico and vitro approaches. Multi-tasking QSAR models were developed using linear discriminant analysis random forest machine learning techniques predicting the responses of interest (G4 interaction, G4 stabilization, selectivity, cytotoxicity) considering variations experimental conditions (e.g., sequences, endpoints, cell lines, buffers, assays). A virtual screening with G4-QuadScreen molecular docking YASARA (AutoDock-Vina) was performed. activities confirmed via FRET melting, FID, viability assays. Validation metrics demonstrated high discriminatory power robustness (the accuracy all is ~>90% training sets ~>80% external sets). evaluations showed that ten screened have capacity to selectively stabilize multiple Three induced strong inhibitory effect various human cancer lines. This pioneering serves accelerate search new leads G4s, reducing false positive outcomes early stages drug discovery. accessible ChemoPredictionSuite website.

Язык: Английский

---