GPO-VAE: modeling explainable gene perturbation responses utilizing GRN-aligned parameter optimization
Bioinformatics · 2025
DOI: 10.1093/bioinformatics/btaf256 ↗Kurz gefasst
Fachartikel in der Zeitschrift Bioinformatics, 2025. Verfasst von Seungheun Baek und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science and Engineering, Korea University, Seoul, 02841, South Korea. Hauptthemen laut MEDLINE: Gene Regulatory Networks, Computational Biology, Deep Learning, Models, Genetic. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Research Foundation of Korea, Ministry of Health&Welfare, Republic of Korea.
Abstract (Originalsprache)
Motivation Predicting cellular responses to genetic perturbations is essential for understanding biological systems and developing targeted therapeutic strategies. While variational autoencoders (VAEs) have shown promise in modeling perturbation responses, their limited explainability poses a significant challenge, as the learned features often lack clear biological meaning. Nevertheless, model explainability is one of the most important aspects in the realm of biological AI. One of the most effective ways to achieve explainability is incorporating the concept of gene regulatory networks (GRNs) in designing deep learning models such as VAEs. GRNs elicit the underlying causal relationships between genes and are capable of explaining the transcriptional responses caused by genetic perturbation treatments. Results We propose GPO-VAE, an explainable VAE enhanced by GRN-aligned Parameter Optimization that explicitly models gene regulatory networks in the latent space. Our key approach is to optimize the learnable parameters related to latent perturbation effects toward GRN-aligned explainability. Experimental results on perturbation prediction show our model achieves state-of-the-art performance in predicting transcriptional responses across multiple benchmark datasets. Furthermore, additional results on evaluating the GRN inference task reveal our model’s ability to generate meaningful GRNs compared to other methods. According to qualitative analysis, GPO-VAE possesses the ability to construct biologically explainable GRNs that align with experimentally validated regulatory pathways. Availability and implementation GPO-VAE is available at https://github.com/dmis-lab/GPO-VAE.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Computer Science and Engineering, Korea University, Seoul, 02841, South Korea.
- LG AI Research, Seoul, 07789, South Korea.
- Department of Biomedical Engineering, Hankuk University of Foreign Studies, Yongin, 17035, South Korea.
- AIGEN Sciences, Seoul, 04778, South Korea.
ORCID-Profile: Jaewoo Kang
Förderung
- National Research Foundation of Korea
- Ministry of Health&Welfare
- Republic of Korea
- ICT Creative Consilience Program
- Institute of Information&Communications Technology Planning&Evaluation
- Korea government
- Korea Health Industry Development Institute
- Ministry of Health&Welfare, Korea
Zitieren
APA
Baek, S., Park, S., Chok, Y. T., Gim, M., & Kang, J. (2025). GPO-VAE: modeling explainable gene perturbation responses utilizing GRN-aligned parameter optimization. Bioinformatics. https://doi.org/10.1093/bioinformatics/btaf256
Vancouver
Baek S, Park S, Chok YT, Gim M, Kang J. GPO-VAE: modeling explainable gene perturbation responses utilizing GRN-aligned parameter optimization. Bioinformatics. 2025. doi:10.1093/bioinformatics/btaf256
BibTeX
@article{Baek2025,
author = {Baek, Seungheun and Park, Soyon and Chok, Yan Ting and Gim, Mogan and Kang, Jaewoo},
title = {GPO-VAE: modeling explainable gene perturbation responses utilizing GRN-aligned parameter optimization},
journal = {Bioinformatics},
year = {2025},
doi = {10.1093/bioinformatics/btaf256}
}
EUVIMED