Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Udit, Udit
Format: Recurso digital
Sprache:
Veröffentlicht: Zenodo 2025
Online-Zugang:https://doi.org/10.5281/zenodo.15764921
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866901855105187840
author Udit, Udit
author_facet Udit, Udit
contents <p><span lang="EN">Long non-coding RNAs (lncRNAs) play diverse and critical roles in gene regulation, chromatin organization, and various cellular processes. Elucidating lncRNA-disease associations is essential for uncovering novel therapeutic targets and understanding disease mechanisms. In this study, we present PrediLnc, a comprehensive platform designed to disseminate predictive insights generated by our model GARNet (Graph convolution Attention RNA Network using stack ensemble to predict LncRNA-disease associations).</span></p> <p><span lang="EN">GARNet integrates multi-level biological information, including lncRNA sequence features, disease ontology terms, associative data at the gene, miRNA, and protein levels, to generate robust predictions. The model architecture leverages autoencoders for dimensionality reduction, graph convolutional networks and self-attention mechanisms for capturing topological and contextual node information, and a stacked ensemble learning framework for final prediction. The ensemble includes five base classifiers: Logistic Regression, Support Vector Machine, Random Forest, Gradient Boosting, and Multi-Layer Perceptron, with a Random Forest serving as the meta-classifier.</span></p> <p><span lang="EN">GARNet achieves strong performance in stratified cross-validation and is available as both a web server and a standalone tool through PrediLnc (</span><span lang="EN"><a href="https://dhanjal-lab.iiitd.edu.in/predilnc.html"><span>https://dhanjal-lab.iiitd.edu.in/predilnc.html</span></a></span><span lang="EN">). The platform's biological relevance has been validated through independent case studies, involving distinct lncRNAs and diseases, with predictions supported by evidence from over 500 PubMed studies, underscoring its reliability for the scientific community.</span></p> <p><strong><span lang="EN">Keywords: </span></strong><span lang="EN">long non-coding RNAs, lncRNA-disease association, disease ontology, deep learning, machine learning, GARNet, PrediLnc</span></p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_15764921
institution Zenodo
language
publishDate 2025
publisher Zenodo
record_format zenodo
spellingShingle PrediLnc: A Platform for disseminating lncRNA-disease associations.
Udit, Udit
<p><span lang="EN">Long non-coding RNAs (lncRNAs) play diverse and critical roles in gene regulation, chromatin organization, and various cellular processes. Elucidating lncRNA-disease associations is essential for uncovering novel therapeutic targets and understanding disease mechanisms. In this study, we present PrediLnc, a comprehensive platform designed to disseminate predictive insights generated by our model GARNet (Graph convolution Attention RNA Network using stack ensemble to predict LncRNA-disease associations).</span></p> <p><span lang="EN">GARNet integrates multi-level biological information, including lncRNA sequence features, disease ontology terms, associative data at the gene, miRNA, and protein levels, to generate robust predictions. The model architecture leverages autoencoders for dimensionality reduction, graph convolutional networks and self-attention mechanisms for capturing topological and contextual node information, and a stacked ensemble learning framework for final prediction. The ensemble includes five base classifiers: Logistic Regression, Support Vector Machine, Random Forest, Gradient Boosting, and Multi-Layer Perceptron, with a Random Forest serving as the meta-classifier.</span></p> <p><span lang="EN">GARNet achieves strong performance in stratified cross-validation and is available as both a web server and a standalone tool through PrediLnc (</span><span lang="EN"><a href="https://dhanjal-lab.iiitd.edu.in/predilnc.html"><span>https://dhanjal-lab.iiitd.edu.in/predilnc.html</span></a></span><span lang="EN">). The platform's biological relevance has been validated through independent case studies, involving distinct lncRNAs and diseases, with predictions supported by evidence from over 500 PubMed studies, underscoring its reliability for the scientific community.</span></p> <p><strong><span lang="EN">Keywords: </span></strong><span lang="EN">long non-coding RNAs, lncRNA-disease association, disease ontology, deep learning, machine learning, GARNet, PrediLnc</span></p>
title PrediLnc: A Platform for disseminating lncRNA-disease associations.
url https://doi.org/10.5281/zenodo.15764921