Identifying Protein-Nucleotide Binding Residues via Grouped Multi-task Learning and Pre-trained Protein Language Models

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

Опубликована: Янв. 9, 2025

The accurate identification of protein-nucleotide binding residues is crucial for protein function annotation and drug discovery. Numerous computational methods have been proposed to predict these residues, achieving remarkable performance. However, due the limited availability high variability nucleotides, predicting diverse nucleotides remains a significant challenge. To address these, we propose NucGMTL, new grouped deep multi-task learning approach designed all observed in BioLiP database. NucGMTL leverages pre-trained language models generate robust sequence embedding incorporates multi-scale along with scale-based self-attention mechanisms capture broader range feature dependencies. effectively harness shared patterns across various utilized distill common representations, taking advantage auxiliary information from similar selected based on task grouping. Performance evaluation benchmark data sets shows that achieves an average area under Precision-Recall curve (AUPRC) 0.594, surpassing other state-of-the-art methods. Further analyses highlight predominant can be reflected by its effective integration models. set source code are freely accessible at: https://github.com/jerry1984Y/NucGMTL.

Язык: Английский

---

Improving Identification of Drug-Target Binding Sites Based on Structures of Targets Using Residual Graph Transformer Network

Biomolecules, Год журнала: 2025, Номер 15(2), С. 221 - 221

Опубликована: Фев. 3, 2025

Improving identification of drug-target binding sites can significantly aid in drug screening and design, thereby accelerating the development process. However, due to challenges such as insufficient fusion multimodal information from targets imbalanced datasets, enhancing performance prediction models remains exceptionally difficult. Leveraging structures targets, we proposed a novel deep learning framework, RGTsite, which employed Residual Graph Transformer Network improve sites. First, residual 1D convolutional neural network (1D-CNN) pre-trained model ProtT5 were extract local global sequence features target, respectively. These then combined with physicochemical properties amino acid residues serve vertex graph. Next, edge incorporated, graph transformer (GTN) was applied more comprehensive features. Finally, fully connected used classify whether site. Experimental results showed that RGTsite outperformed existing state-of-the-art methods key evaluation metrics, F1-score (F1) Matthews Correlation Coefficient (MCC), across multiple benchmark datasets. Additionally, conducted interpretability analysis for through real-world cases, confirmed effectively identify practical applications.

Язык: Английский

---

Machine learning approaches for predicting protein-ligand binding sites from sequence data

Frontiers in Bioinformatics, Год журнала: 2025, Номер 5

Опубликована: Фев. 3, 2025

Proteins, composed of amino acids, are crucial for a wide range biological functions. Proteins have various interaction sites, one which is the protein-ligand binding site, essential molecular interactions and biochemical reactions. These sites enable proteins to bind with other molecules, facilitating key Accurate prediction these pivotal in computational drug discovery, helping identify therapeutic targets facilitate treatment development. Machine learning has made significant contributions this field by improving interactions. This paper reviews studies that use machine predict from sequence data, focusing on recent advancements. The review examines embedding methods architectures, addressing current challenges ongoing debates field. Additionally, research gaps existing literature highlighted, potential future directions advancing discussed. study provides thorough overview sequence-based approaches predicting offering insights into state possibilities.

Язык: Английский

---

HcTRET1 is critical for epidermal chitin synthesis in Hyphantria cunea

Insect Molecular Biology, Год журнала: 2025, Номер unknown

Опубликована: Май 9, 2025

Abstract In insects, trehalose is critical for growth and development, as well environmental stress response, which mainly transported by transporters (TRETs). Over nearly two decades, the physiological functions of TRETs in insect growth, reproduction response have been elucidated. However, role chitin synthesis remains not fully understood. Here, we identified HcTRET1 gene from Hyphantria cunea , a major Lepidoptera pest agriculture forestry. The especially synthesis, was discussed dsRNA‐mediated knockdown. Bioassay showed that knockdown did affect larval development survival H. but it significantly reduced pupa formation rate. Additionally, silencing increased levels fat body decreased them hemolymph, suggesting plays key homeostasis. Moreover, downregulated genes ( HcGFAT HcUAP HcCHSA ), resulting remarkable reduction content epidermis. larvae at 42°C. Taken together, these results demonstrated played larva–pupa transition, vivo homeostasis, epidermal biosynthesis larvae. parallel, its important function to high‐temperature has verified well. findings expand our understanding TRET1 providing new perspective regulate biosynthesis.

Язык: Английский

---

Exploring the phytochemical profile and therapeutic investigations on Moringa concanensis Nimmo pod husk extracts: An integrated in vitro and in silico approach

Biocatalysis and Agricultural Biotechnology, Год журнала: 2024, Номер 58, С. 103234 - 103234

Опубликована: Май 16, 2024

Язык: Английский

---

A Point Cloud Graph Neural Network for Protein–Ligand Binding Site Prediction

International Journal of Molecular Sciences, Год журнала: 2024, Номер 25(17), С. 9280 - 9280

Опубликована: Авг. 27, 2024

Predicting protein-ligand binding sites is an integral part of structural biology and drug design. A comprehensive understanding these essential for advancing innovation, elucidating mechanisms biological function, exploring the nature disease. However, accurately identifying remains a challenging task. To address this, we propose PGpocket, geometric deep learning-based framework to improve site prediction. Initially, protein surface converted into point cloud, then chemical properties each are calculated. Subsequently, cloud graph constructed based on inter-point distances, neural network (GNN) applied extract analyze information predict potential sites. PGpocket trained scPDB dataset, its performance verified two independent test sets, Coach420 HOLO4K. The results show that achieves 58% success rate dataset 56% HOLO4K dataset. These surpass competing algorithms, demonstrating PGpocket's advancement practicality

Язык: Английский

---

Next-Generation Carbazole-Linked 1,2,4-Triazole-Thione Derivatives: Strategic Design, Synthesis, Molecular Docking, and Evaluation of Antidiabetic Potential

ACS Omega, Год журнала: 2024, Номер 10(1), С. 848 - 861

Опубликована: Дек. 25, 2024

Currently, available therapies for diabetes cannot achieve normal sugar values in a high percentage of treated patients. This work synthesized series carbazole-triazole-thione derivatives, and their potential antidiabetic activity was investigated against the key diabetic enzymes α-amylase glycosidase. Normal human hepatic stellate cells (LX-2) were employed to assess cytotoxicity safety, followed by vivo testing investigate hypoglycemic effect most promising agent. As result, set 18 carbazole-1,2,4-triazole-thione derivatives synthesized. Seven structures demonstrated inhibitory enzyme, with IC50 lower than 6.4 μM. Among them, compounds C5f, C5o, C5r exhibited highest potency, 0.56, 0.53, 0.97 μM, respectively, compared well-known inhibitor acarbose, which has an value 5.31 Exploring inhibition potency these α-glucosidase enzyme revealed that C5f act as moderate inhibitors, 11.03 13.76 respectively. Moreover, at 100 μM concentration, evaluated showed negligible cytotoxic LX-2 cell lines, particularly C5o C5s, 3-fold positive control 5-Flururicle (cell viability 13.45%). Thus, compound selected evaluation, after administering five doses this (10 mg/kg) group III mice, significant reduction glucose concentration observed, bringing it down from 290.54 216.15 mg/dL, comparison did not show blood level. These observed vitro results upheld performing chemoinformatic studies elucidated binding interactions active within enzyme's site highlighted critical roles both 1,2,4-triazole-3-thione carbazole scaffolds interactions. Finally, drug-likeness profiles our suggest candidates further clinical trials.

Язык: Английский

---

Computational Tools for Structural Analysis of Proteins

Elsevier eBooks, Год журнала: 2024, Номер unknown

Опубликована: Янв. 1, 2024

Язык: Английский

---

Varied sensitivity to boscalid among different Clarireedia species causing dollar spot in turfgrass

Pesticide Biochemistry and Physiology, Год журнала: 2024, Номер 204, С. 106029 - 106029

Опубликована: Июль 18, 2024

Язык: Английский

---

Structural Genomics

Elsevier eBooks, Год журнала: 2024, Номер unknown

Опубликована: Янв. 1, 2024

---