Class-balanced negative training sets for improving classifier model predictions of enhancer-promoter interactions
Our 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 .
Authors
- Osamu Maruyama (ORCID: https://orcid.org/0000-0002-5760-1507)
- Tsukasa Koga (ORCID: https://orcid.org/0009-0005-7963-7200)
Institutions
- Kyushu University (JP)
- Minami Kyushu University (JP)
Publication Details
- Journal
- BMC Bioinformatics
- Published
- 2025-06-02
- DOI
- https://doi.org/10.1186/s12859-025-06171-8
- Primary Topic
- Genomics and Chromatin Dynamics
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Japan Society for the Promotion of Science