Model Agnostic Semi-Supervised Meta-Learning Elucidates Understudied Out-of-distribution Molecular Interactions DOI Creative Commons
You Wu, Li Xie, Yang Liu

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: May 20, 2023

Many biological problems are understudied due to experimental limitations and human biases. Although deep learning is promising in accelerating scientific discovery, its power compromises when applied with scarcely labeled data distribution shifts. We developed a semi-supervised meta framework - Meta Model Agnostic Pseudo Label Learning (MMAPLE) address these challenges by effectively exploring out-of-distribution (OOD) unlabeled transfer fails. The of MMAPLE demonstrated multiple applications: predicting OOD drug-target interactions, hidden metabolite-enzyme interspecies microbiome metabolite-human receptor where chemicals or proteins unseen dramatically different from those training data. achieves 11% 242% improvement the prediction-recall on benchmarks over baseline models. Using MMAPLE, we reveal novel metabolite-protein interactions that validated bioactivity assays fill missing links microbiome-human interactions. general explore previously unrecognized domains beyond reach present computational techniques.

Language: Английский

Chloride ions in health and disease DOI Open Access
Satish K. Raut,

K.G Singh,

Shridhar Sanghvi

et al.

Bioscience Reports, Journal Year: 2024, Volume and Issue: 44(5)

Published: April 4, 2024

Chloride is a key anion involved in cellular physiology by regulating its homeostasis and rheostatic processes. Changes Cl- concentration result differential regulation of functions such as transcription translation, post-translation modifications, cell cycle proliferation, volume, pH levels. In intracellular compartments, modulates the function lysosomes, mitochondria, endosomes, phagosomes, nucleus, endoplasmic reticulum. extracellular fluid (ECF), present blood/plasma interstitial compartments. A reduction levels ECF can volume contraction. physiological principal compensatory ion for movement major cations Na+, K+, Ca2+. Over past 25 years, we have increased our understanding signaling mediated Cl-, which has helped molecular metabolic changes observed pathologies with altered Here, review various organs channels responsible transportation, recent information on roles.

Language: Английский

Citations

9
Microbial molecules, metabolites, and malignancy DOI Creative Commons
Ryan M. Thomas

Neoplasia, Journal Year: 2025, Volume and Issue: 60, P. 101128 - 101128

Published: Jan. 18, 2025

Language: Английский

Citations

0
Using Zebrafish to Study the Mechanisms That Underlie Down Syndrome DOI
Anna J. Moyer, Summer B. Thyme

Published: Jan. 1, 2025

Language: Английский

Citations

0
Semi-supervised meta-learning elucidates understudied molecular interactions DOI Creative Commons
You Wu, Li Xie, Yang Liu

et al.

Communications Biology, Journal Year: 2024, Volume and Issue: 7(1)

Published: Sept. 9, 2024

Language: Английский

Citations

3
ZBTB11 Depletion Targets Metabolic Vulnerabilities in K-Ras Inhibitor Resistant PDAC DOI Creative Commons
Nathan L. Tran, Jiewei Jiang,

Min Ma

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: May 21, 2024

Over 95% of pancreatic ductal adenocarcinomas (PDAC) harbor oncogenic mutations in K-Ras. Upon treatment with K-Ras inhibitors, PDAC cancer cells undergo metabolic reprogramming towards an oxidative phosphorylation-dependent, drug-resistant state. However, direct inhibition complex I is poorly tolerated patients due to on-target induction peripheral neuropathy. In this work, we develop molecular glue degraders against ZBTB11, a C

Language: Английский

Citations

2
Chemical proteomics accelerates the target discovery of natural products DOI
Shujie He, Jun Li, Jinming Zhou

et al.

Biochemical Pharmacology, Journal Year: 2024, Volume and Issue: 230, P. 116609 - 116609

Published: Nov. 5, 2024

Language: Английский

Citations

2
Best practices for managing and disseminating resources and outreach and evaluating the impact of the IDG Consortium DOI Creative Commons
D. Vidović, Anna Waller, Jayme Holmes

et al.

Drug Discovery Today, Journal Year: 2024, Volume and Issue: 29(5), P. 103953 - 103953

Published: March 18, 2024

The Illuminating the Druggable Genome (IDG) consortium generated reagents, biological model systems, data, informatic databases, and computational tools. Resource Dissemination Outreach Center (RDOC) played a central administrative role, organized internal meetings, fostered collaboration, coordinated consortium-wide efforts. RDOC developed deployed Management System (RMS) to enable efficient workflows for collecting, accessing, validating, registering, publishing resource metadata. IDG policies repositories standardized representations of resources were established, adopting FAIR (findable, accessible, interoperable, reusable) principles. also metrics impact. initiatives included digital content, Protein Illumination Timeline (representing milestones in generating data reagents), Target Watch publication series, e-IDG Symposium leveraging social media platforms.

Language: Английский

Citations

1
Endolysosomal processing of neuron-derived signaling lipids regulates autophagy and lipid droplet degradation in astrocytes DOI Creative Commons
Jagannatham Naidu Bhupana,

A. Pabon,

Ho Hang Leung

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 9, 2024

Abstract Astrocytes support brain metabolism by processing, storing, and appropriating metabolites. Dynamic regulation of metabolic activities in astrocytes is critical to meeting the demands other cells. During neuronal stress, lipid metabolites are transferred from neurons astrocytes, where they stored droplets (LDs). However, it not clear whether how neuron-derived lipids trigger adaptation astrocytes. Here, we uncover an endolysosomal function that mediates a neuron-astrocyte transcellular signaling paradigm. We identify Tweety homolog 1 (TTYH1) as astrocyte-enriched transmembrane protein localized endolysosomes, facilitates autophagic flux droplet (LD) degradation. Astrocyte-specific deletion Ttyh1 mice loss TTYH1 ortholog Drosophila lead accumulation neutral lipids. Computational experimental evidence suggests clearance ceramide 1-phosphate (C1P), sphingolipid dampens LD breakdown mouse human found inflammatory cytokine IL-1β induces upregulation C1P biosynthesis. Concurrently, secreted cause impairment Whereas deficiency exacerbates catabolic blockage, inhibiting synthesis restores normalizes contents Thus, rely on mitigate effects Taken together, our findings reveal neuron-initiated paradigm culminates

Language: Английский

Citations

0
Unlocking precision medicine: Innovative strategies for druggable target identification and therapeutic enhancement DOI Creative Commons

Yang Liao,

Zhangle Wei,

Hangwei Xu

et al.

Precision medication., Journal Year: 2024, Volume and Issue: 1(1), P. 16 - 29

Published: Nov. 1, 2024

Language: Английский

Citations

0
Model Agnostic Semi-Supervised Meta-Learning Elucidates Understudied Out-of-distribution Molecular Interactions DOI Creative Commons
You Wu, Li Xie, Yang Liu

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: May 20, 2023

Many biological problems are understudied due to experimental limitations and human biases. Although deep learning is promising in accelerating scientific discovery, its power compromises when applied with scarcely labeled data distribution shifts. We developed a semi-supervised meta framework - Meta Model Agnostic Pseudo Label Learning (MMAPLE) address these challenges by effectively exploring out-of-distribution (OOD) unlabeled transfer fails. The of MMAPLE demonstrated multiple applications: predicting OOD drug-target interactions, hidden metabolite-enzyme interspecies microbiome metabolite-human receptor where chemicals or proteins unseen dramatically different from those training data. achieves 11% 242% improvement the prediction-recall on benchmarks over baseline models. Using MMAPLE, we reveal novel metabolite-protein interactions that validated bioactivity assays fill missing links microbiome-human interactions. general explore previously unrecognized domains beyond reach present computational techniques.

Language: Английский

Citations

0