Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

DeePROG: Deep Attention-Based Model for Diseased Gene Prognosis by Fusing Multi-Omics Data.

Pratik Dutta, Aditya Prakash Patra, Sriparna Saha

    IEEE/ACM Transactions on Computational Biology and Bioinformatics
    |June 24, 2021
    PubMed
    Summary

    DeePROG, a novel deep multi-modal model, accurately predicts disease-affected genes using diverse omics data. This approach enhances understanding of gene functions for drug design and personalized medicine.

    Related Concept Videos

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    TSProm: deep learning framework to predict tissue-specific regulatory logic.

    NAR genomics and bioinformatics·2026
    Same author

    HViLM: A Foundation Model for Viral Genomics Enables Multi-Task Prediction of Pathogenicity, Transmissibility, and Host Tropism.

    bioRxiv : the preprint server for biology·2026
    Same author

    Enhancing adverse drug event extraction and summarization for cancer drugs through large language models.

    Journal of biomedical informatics·2026
    Same author

    A DNABERT based deep learning framework for predicting transcription factor binding sites.

    Scientific reports·2026
    Same author

    TSProm: Deciphering the Genomic Context of Tissue Specificity.

    bioRxiv : the preprint server for biology·2025
    Same author

    The Role of Place in Fostering Belonging and Science Identity Development for Incoming Ecology and Evolutionary Biology Graduate Students: Perspectives From a Two-Year Program Evaluation.

    Ecology and evolution·2025

    Area of Science:

    • Genomics
    • Bioinformatics
    • Computational Biology

    Background:

    • Understanding gene functions is critical for disease pathobiology, drug design, personalized medicine, and diagnostics.
    • Heterogeneous genomic datasets from omics technologies (gene expression, DNA sequence, 3D structures) offer valuable insights.
    • Interpreting complex gene functions from these diverse data sources remains a significant challenge.

    Purpose of the Study:

    • To propose a novel self-attention based deep multi-modal model, DeePROG, for disease-affected gene prognosis.
    • To leverage heterogeneous omics data, including gene expression, DNA sequence, and 3D protein structures, for improved gene function prediction.
    • To develop and evaluate advanced deep learning architectures for biomedical data analysis.

    Main Methods:

    Related Experiment Videos

    • Developed DeePROG, a self-attention based deep multi-modal model integrating gene expression, DNA sequence, and 3D protein structure data.
    • Utilized context-specific deep learning models for feature extraction from each data modality.
    • Implemented attention-based deep bi-modal architectures alongside DeePROG for enhanced prognosis.
    • Assessed model performance using Computational Assessment of Function Annotation (CAFA2) metrics, ROC curves, and Welch's t-test.

    Main Results:

    • DeePROG significantly outperformed baseline models in predicting disease-affected genes.
    • The model demonstrated robust performance across various performance metrics, including in high-class imbalance settings.
    • Analysis using ROC curves and statistical significance tests validated the efficacy of DeePROG.

    Conclusions:

    • The proposed DeePROG model offers a powerful approach for gene prognosis using multi-modal omics data.
    • This methodology advances the understanding of gene functions in disease pathobiology.
    • The findings support the application of deep learning in personalized medicine and next-generation diagnostics.