• Home
  • Search
  • Class-balanced negative training sets for improving classifier model predictions of enhancer-promoter interactions
  • https://doi.org/10.1186/s12859-025-06171-8Copy DOI Icon

Class-balanced negative training sets for improving classifier model predictions of enhancer-promoter interactions

Show More
  • Abstract
  • Literature Map
  • References
  • Similar Papers
Abstract

BackgroundEnhancers regulate gene expression by forming DNA loops, thereby bringing themselves in close proximity to the target gene promoter. The human genome contains hundreds of thousands of enhancers, vastly outnumbering its 20,000–25,000 protein-coding genes, highlighting the importance of enhancer-promoter interactions (EPIs) in gene regulation. Supervised learning models have been developed to predict EPIs, often using experimentally validated interacting enhancer-promoter pairs and artificially generated negative samples. However, the lack of reliable negative samples presents a challenge. Current methods randomly select pairs from unlabeled data, leading to class imbalance and reduced predictive performance. This imbalance, where enhancers and promoters are unevenly distributed between the positive and negative sets, hinders classifiers from learning meaningful patterns. Therefore, constructing more reliable negative samples is crucial for improving the accuracy of EPI predictions.ResultsWe developed two methods to generate class-balanced negative training sets for EPI classifiers: one based on maximum flow and the other on Gibbs sampling. We evaluated these methods with the TargetFinder and TransEPI classifiers across five and six cell lines, respectively. The trained models were tested using a common negative test set. Our negative training sets significantly improved the prediction performance across several metrics, including precision, recall, and area under the receiver operating characteristic curve.ConclusionsOur findings demonstrate that carefully designed negative samples can enhance the performance of EPI classifiers. Further advanced methods in generating negative EPIs should further improve prediction accuracy. The source code is available at https://github.com/maruyama-lab-design/CBOEP2.

Similar Papers
  • Preprint Article

Class-balanced negative training sets for improving classifier model predictions of enhancer-promoter interactions

  • Apr 24, 2025
  • Research Square
  • Osamu Maruyama +1
  • PDF
  • Research Article
  • Citations4

Predicting potential miRNA-disease associations based on more reliable negative sample selection

  • Oct 17, 2022
  • BMC Bioinformatics
  • Ruiyu Guo +4
  • Research Article
  • Citations4

SURE: Screening unlabeled samples for reliable negative samples based on reinforcement learning

  • Feb 03, 2023
  • Information Sciences
  • Ying Li +7
  • Research Article
  • Citations32

MGCNSS: miRNA-disease association prediction with multi-layer graph convolution and distance-based negative sample selection strategy.

  • Mar 27, 2024
  • Briefings in bioinformatics
  • Zhen Tian +4
  • Research Article
  • Citations361

Improving compound-protein interaction prediction by building up highly credible negative samples.

  • Jun 10, 2015
  • Bioinformatics
  • Hui Liu +4
  • Research Article
  • Citations5

Drug-target interaction prediction using reliable negative samples and effective feature selection methods

  • Jun 20, 2022
  • Journal of Pharmacological and Toxicological Methods
  • Mohammad Morovvati Sharifabad +2
  • Research Article
  • Citations14

An ensemble learning framework for potential miRNA-disease association prediction with positive-unlabeled data

  • Aug 24, 2021
  • Computational Biology and Chemistry
  • Yao Wu +3
  • Research Article
  • Citations6

FCCCSR_Glu: a semi-supervised learning model based on FCCCSR algorithm for prediction of glutarylation sites.

  • Sep 27, 2022
  • Briefings in Bioinformatics
  • Qiao Ning +4
  • Research Article
  • Citations6

SMDPG: Identifying Metabolite-Disease Associations via Optimized Negative Sampling and Sparse Graph Convolutional Network.

  • Mar 01, 2025
  • IEEE transactions on computational biology and bioinformatics
  • Yiran Huang +4
  • PDF
  • Research Article
  • Citations2

Multi-Class Positive and Unlabeled Learning for High Dimensional Data Based on Outlier Detection in a Low Dimensional Embedding Space

  • Sep 05, 2022
  • Electronics
  • Cheong Hee Park
  • Research Article
  • Citations6

Ensembles of density estimators for positive-unlabeled learning

  • Mar 01, 2019
  • Journal of Intelligent Information Systems
  • T M A Basile +4
  • Book Chapter
  • Citations7

Density Estimators for Positive-Unlabeled Learning

  • Jan 01, 2018
  • Teresa M A Basile +4
  • Research Article
  • Citations38

An integrated framework for the identification of potential miRNA-disease association based on novel negative samples extraction strategy

  • Jan 28, 2019
  • RNA Biology
  • Chun-Chun Wang +3
  • Research Article
  • Citations40

Inverse similarity and reliable negative samples for drug side-effect prediction

  • Feb 01, 2019
  • BMC Bioinformatics
  • Yi Zheng +4
  • Research Article

Cluster-guided adversarial graph contrastive learning.

  • Jul 04, 2026
  • Neural networks : the official journal of the International Neural Network Society
  • Dong Huang +4
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.