• HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
Wednesday, November 18, 2026
BIOENGINEER.ORG
No Result
View All Result
  • Login
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
No Result
View All Result
Bioengineer.org
No Result
View All Result
Home NEWS Science News Biology

Ontology Pre-Training Accelerates Deep Learning Predictions for Biological Metabolite Properties

by
November 17, 2026
in Biology
Reading Time: 2 mins read
0
Ontology Pre-Training Accelerates Deep Learning Predictions for Biological Metabolite Properties

Ontology Pre-Training Accelerates Deep Learning Predictions for Biological Metabolite Properties

Share on FacebookShare on TwitterShare on LinkedinShare on RedditShare on Telegram

Researchers have developed an intermediate training approach called ontology pre-training that integrates structured biological knowledge into deep learning models to improve predictions of metabolite behaviors. Recent computational advances show that integrating expert domain knowledge enhances predictive performance in complex biological settings, where curated bio-ontologies offer rich hierarchies of biological classifications.

The generalisable methodology embeds classification hierarchies derived directly from bio-ontologies into deep neural networks. Functioning between general-purpose pre-training and task-specific fine-tuning, this intermediate step injects curated expert representations before models adapt to specialized downstream targets. Researchers Charlotte Tumescheit, Martin Glauer, Simon Flügel, Martin Larralde, Fabian Neuhaus, Till Mossakowski, and Janna Hastings evaluated the pipeline across datasets from MoleculeNet.

Testing demonstrated improved predictive accuracy alongside reduced training times across a wide range of prediction tasks concerning metabolite activities, functions, and properties in living systems. The model achieved its largest performance gains across regression benchmarks, notably in predicting the aqueous solubility and lipophilicity of chemical molecules, while demonstrating robust improvements across most classification tasks.

The research team confirmed that this ontology pre-training strategy provides an intuitive framework that can be adapted to support diverse external knowledge sources, underlying model architectures, and biological prediction tasks. By linking hierarchical expert structures directly to deep learning pipelines, the method addresses longstanding predictive challenges in biological computation.

Subject of Research: Embedding bio-ontology classification hierarchies into deep learning models for metabolite prediction

Article Title: Ontology pre-training improves machine learning-based predictions for metabolites

Article References: Tumescheit, C., Glauer, M., Flügel, S., Larralde, M., Neuhaus, F., Mossakowski, T., & Hastings, J. (2026). Ontology pre-training improves machine learning-based predictions for metabolites. PLOS Computational Biology, 22(11), e1014925. https://doi.org/10.1371/journal.pcbi.1014925

Image Credits: AI Generated

DOI: 10.1371/journal.pcbi.1014925

Keywords: bio-ontologies, ontology pre-training, deep learning, machine learning, metabolites, predictive performance, training time reduction, MoleculeNet, regression tasks, classification tasks, lipophilicity prediction, aqueous solubility, expert knowledge integration, living systems, biological problems, task-specific fine-tuning

News Source: Blake Davidson. (November 17, 2026). Ontology Pre-Training Accelerates Deep Learning Predictions for Biological Metabolite Properties. Scienmag.

Tags: aqueous solubilitybio-ontologiesbiological problemsClassification Tasksdeep learningexpert knowledge integrationlipophilicity predictionliving systemsMachine LearningmetabolitesMoleculeNetontology pre-trainingpredictive performanceregression taskstask-specific fine-tuningtraining time reduction
Share12Tweet7Share2ShareShareShare1

Related Posts

Endothelial Subtypes Show Contrasting Hypoxia Signals in Engineered 3D Hydrogels

Endothelial Subtypes Show Contrasting Hypoxia Signals in Engineered 3D Hydrogels

November 17, 2026
RSV Activity and Interference Shape Human Metapneumovirus Timing in US Study

RSV Activity and Interference Shape Human Metapneumovirus Timing in US Study

November 17, 2026

Recent Speciation in Ancient Tropical Fern Lineages Linked to Andes Uplift

November 17, 2026

Dynamic Segregated Model Predicts CHO Cell Cycle and Metabolism in Perfusion Cultures

November 17, 2026

POPULAR NEWS

  • Survey Reveals Lack of Parental Awareness Regarding Sibling Visitation Policies in PICUs

    Survey Reveals Lack of Parental Awareness Regarding Sibling Visitation Policies in PICUs

    29 shares
    Share 12 Tweet 7
  • Review Details How Circulating Tumor Cells Survive Bloodstream Stresses to Drive Metastasis

    29 shares
    Share 12 Tweet 7
  • Laparoscopic Colectomy Cuts Surgical Site Infection Risk in Obese Adults

    29 shares
    Share 12 Tweet 7
  • Aging-Related Immune Cell Imbalances Linked to Subclinical Coronary Artery Disease

    29 shares
    Share 12 Tweet 7

About

We bring you the latest biotechnology news from best research centers and universities around the world. Check our website.

Follow us

Recent News

Intraoral Scanners Achieve High In Vitro Accuracy for Neonatal Cleft Models

Italian Short Form Hip Score Proves Valid After Total Hip Arthroplasty

Case Report Details Inherited ATM Gene Mutation in Aggressive Familial Prostate Cancer

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 88 other subscribers
  • Contact Us

Bioengineer.org © Copyright 2023 All Rights Reserved.

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • Homepages
    • Home Page 1
    • Home Page 2
  • News
  • National
  • Business
  • Health
  • Lifestyle
  • Science

Bioengineer.org © Copyright 2023 All Rights Reserved.