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

miTAR: a hybrid deep learning-based approach for predicting miRNA targets.

Tongjun Gu1,2, Xiwu Zhao3, William Bradley Barbazuk4,5,6

  • 1Bioinformatics, Interdisciplinary Center for Biotechnology Research, University of Florida, Gainesville, FL, USA. tgu@ufl.edu.

BMC Bioinformatics
|February 28, 2021
PubMed
Summary

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

Nascent Pre-Platelets and B4GALT1 Glycosylation Contribute to 5-FU-induced Bone Marrow Recovery.

Blood advances·2026
Same author

SurvMarker: an R package for identifying survival-associated molecular features using PCA-based weighted scores.

BMC bioinformatics·2026
Same author

Defective cerebrovascular development in mice lacking TFPI is restored by activated protein C.

Blood·2026
Same author

Platelet glycosylation in myelodysplastic syndromes correlates with disease severity.

Blood advances·2025
Same author

Brain 5-hydroxymethylcytosine alterations are associated with Alzheimer's disease neuropathology.

Nature communications·2025
Same author

Prothrombotic antibodies targeting the spike protein's receptor-binding domain in severe COVID-19.

Blood·2024

A new hybrid deep learning approach accurately predicts microRNA (miRNA) targets using raw gene sequences. This method integrates CNNs and RNNs, outperforming existing tools and offering a robust solution for miRNA target identification.

Area of Science:

  • Computational biology
  • Genomics
  • Bioinformatics

Background:

  • microRNAs (miRNAs) regulate gene expression and are crucial in biological processes.
  • Existing miRNA target prediction methods rely on complex, pre-defined features, limiting accuracy and efficiency.
  • There is a need for more accurate and resource-efficient computational approaches for miRNA target identification.

Purpose of the Study:

  • To develop a novel hybrid deep learning (DL) approach for highly accurate miRNA target prediction.
  • To integrate Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) for enhanced feature learning.
  • To provide an effortless and robust tool for identifying miRNA targets from raw sequence data.

Main Methods:

  • Developed a hybrid DL model combining CNNs for spatial feature extraction and RNNs for sequential feature learning.
Keywords:
Convolutional neural networksDeep learningHybrid modelMiRNA targetRecurrent neural networks

Related Experiment Videos

  • Utilized raw miRNA and gene sequences as direct inputs, eliminating the need for manual feature engineering.
  • Trained and evaluated models on human datasets, comparing performance against existing state-of-the-art methods.

    • Main Results:

    • The proposed DL-based approach significantly outperforms previous methods in miRNA target prediction accuracy on test datasets.
    • The hybrid model demonstrates superior performance on independent datasets and robustness on smaller datasets.
    • Incorporating a Max Pooling layer between CNN and RNN components further enhances model performance and generalization.
    • A unified model was developed, proving robust across different input datasets.

    Conclusions:

    • The novel DL-based approach offers a significant advancement in miRNA target prediction accuracy.
    • The developed tool, miTAR, provides an accessible and effective solution for researchers.
    • Max Pooling integration is beneficial for hybrid DL models, potentially mitigating overfitting in miRNA target prediction.