Nature reviews. Cancer, Год журнала: 2025, Номер unknown
Опубликована: Янв. 14, 2025
Язык: Английский
Процитировано
2Towards a more accurate and reliable evaluation of machine learning protein–protein interaction prediction model performance in the presence of unavoidable dataset biases
Berichte aus der medizinischen Informatik und Bioinformatik/Journal of integrative bioinformatics, Год журнала: 2025, Номер unknown
Опубликована: Апрель 1, 2025
Abstract The characterization of protein-protein interactions (PPIs) is fundamental to understand cellular functions. Although machine learning methods in this task have historically reported prediction accuracies up 95 %, including those only using raw protein sequences, it has been highlighted that could be overestimated due the use random splits and metrics do not take into account potential biases datasets. Here, we propose a per-protein utility metric, pp_MCC, able show drop performance both unseen-protein scenarios. We tested ML models based on sequence embeddings. pp_MCC metric evidences reduced even split, reaching levels similar shown by MCC computed over an unseen drops further when used split scenario. Thus, give more realistic estimation while allowing splits, which interesting for protein-centric studies. Given low adjusted obtained, there seems room improvement primary information, suggesting need inclusion complementary data, accompanied with metric.
Язык: Английский
Процитировано
0Artificial intelligence in peptide-based drug design
Drug Discovery Today, Год журнала: 2025, Номер unknown, С. 104300 - 104300
Опубликована: Янв. 1, 2025
Язык: Английский
Процитировано
0A Deep Learning Method for Predicting Interactions for Intrinsically Disordered Regions of Proteins
Опубликована: Янв. 1, 2025
Язык: Английский
Процитировано
0Targeting cell cycle arrest in breast cancer by phytochemicals from Caryto urens L. fruit ethyl acetate fraction: in silico and in vitro validation
Journal of Ayurveda and Integrative Medicine, Год журнала: 2025, Номер 16(2), С. 101095 - 101095
Опубликована: Март 1, 2025
Caryota urens, also known as Shivjata, has been documented in ancient Indian texts for its therapeutic benefits, addressing conditions from seminal weakness to gastric ulcers. This study aims investigate contemporary medicinal potential treating breast cancer. The focuses on exploring the of urens fruit against cancer, specifically targeting cell cycle genes CDK1, CDC25A, and PLK1 through bioinformatics, network pharmacology, vitro validation. Using mass spectrometry nuclear magnetic resonance (NMR), 60 key phytoconstituents were identified. Bioinformatics analysis, integrating Gene Cards GEO databases, 15,474 cancer-associated focusing HR+/HER2-subtype Molecular docking qPCR validated interactions phytoconstituents, particularly Episesamin, with PLK1. In studies conducted MCF7 line, supplemented by ROC survival analyses evaluate diagnostic potential. bioinformatics analysis identified pivotal regulating progression cancer tumorigenesis. Network pharmacology indicated that especially downregulated these cells. confirmed interactions, underscored their significance. suggests extract, may inhibit metastasis downregulating PLK1, offering promising new strategies emphasizing value experimental methods research.
Язык: Английский
Процитировано
0Target-Specific De Novo Peptide Binder Design with DiffPepBuilder
Journal of Chemical Information and Modeling, Год журнала: 2024, Номер unknown
Опубликована: Сен. 12, 2024
Despite the exciting progress in target-specific
Язык: Английский
Процитировано
2Understanding the Cryptosporidium species and their challenges to animal health and livestock species for informed development of new, specific treatment strategies
Frontiers in Parasitology, Год журнала: 2024, Номер 3
Опубликована: Авг. 6, 2024
Cryptosporidium species are parasitic organisms of vertebrates with a worldwide distribution. They have an important impact globally upon human and animal health, livestock productivity. The life cycle these is complex difficult to disrupt improve food security economic growth. This may contribute the fact that no new treatment strategy has been widely accepted or applied in for years. Here we consider natural history parasites, their biochemistry impact. Using recent developments understanding parasites then viable affordable approaches enhancing control effects on livestock. These based advances drug discovery, omics research artificial intelligence applications veterinary medicine indicate putative therapeutic approaches.
Язык: Английский
Процитировано
0Novel Gurmarin-like Peptides from Gymnema sylvestre and their Interactions with the Sweet Taste Receptor T1R2/T1R3
Chemical Senses, Год журнала: 2024, Номер 49
Опубликована: Янв. 1, 2024
Gymnema sylvestre (GS) is a traditional medicinal plant known for its hypoglycemic and hypolipidemic effects. Gurmarin (hereafter Gur-1) the only active peptide in GS. Gur-1 has suppressive sweet taste effect rodents but no or very weak humans. Here, 8 gurmarin-like peptides (Gur-2 to Gur-9) their isoforms are reported GS transcriptome. The molecular mechanism of suppression by still largely unknown. Therefore, complete architecture human mouse receptors T1R2/T1R3 interaction with Gur-9 were predicted AlphaFold-Multimer (AF-M) validated. Only Gur-2 interact receptor. Indeed, bind region cysteine-rich domain (CRD) transmembrane (TMD) T1R2 subunit. In contrast, binds TMD This result suggests that may have Furthermore, AF-M Gα-gustducin, protein involved transduction, interacts intracellular These results highlight an unexpected diversity provide receptor putative binding sites Gur-1, Gur-2, Gα-gustducin. addition, serve as promising drug scaffolds development antidiabetic molecules.
Язык: Английский
Процитировано
0Prediction of protein interactions between pine and pine wood nematode using deep learning and multi-dimensional feature fusion
Frontiers in Plant Science, Год журнала: 2024, Номер 15
Опубликована: Дек. 2, 2024
Pine Wilt Disease (PWD) is a devastating forest disease that has serious impact on ecological balance ecological. Since the identification of plant-pathogen protein interactions (PPIs) critical step in understanding pathogenic system pine wilt disease, this study proposes Multi-feature Fusion Graph Attention Convolution (MFGAC-PPI) for predicting PPIs based deep learning. Compared with methods single-feature information, MFGAC-PPI obtains more 3D characterization information by utilizing AlphaFold and combining sequence features to extract multi-dimensional via Transform improved GCN. The performance was compared current representative sequence-based, structure-based hybrid characterization, demonstrating its superiority across all metrics. experiments showed learning feature effectively ability plant pathogen PPI prediction tasks. Meanwhile, network consisting 2,688 interacting pairs constructed MFGAC-PPI, which made it possible systematically discover new resistance genes trees promoted interactions.
Язык: Английский
Процитировано
0A deep learning method for predicting interactions for intrinsically disordered regions of proteins
bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown
Опубликована: Дек. 20, 2024
Intrinsically disordered proteins or regions (IDPs IDRs) exist as ensembles of conformations in the monomeric state and can adopt diverse binding modes, making their experimental computational characterization challenging. Here, we developed Disobind, a deep-learning method that predicts inter-protein contact maps interface residues for an IDR partner protein, leveraging sequence embeddings from protein language model. Several current methods, contrast, provide partner-independent predictions, require structure either and/or are limited by MSA quality. Disobind performs better than AlphaFold-multimer AlphaFold3. Combining predictions further improves performance. However, is to binary IDP-partner complexes, where two known bind, input fragments less one hundred long. The be used localize IDRs integrative structures large assemblies, characterize protein-protein interactions involving IDRs, modulate IDR-mediated interactions.
Язык: Английский
Процитировано
0