ST-ConMa: a multimodal foundation framework for spatial transcriptomics via image–gene contrastive and matching learning
Briefings in Bioinformatics · 01.09.2026 · 27 (5)
DOI: 10.1093/bib/bbag535 ↗Kurz gefasst
Fachartikel in der Zeitschrift Briefings in Bioinformatics, online erschienen am 28.09.2026. Verfasst von Taelim Yeom und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of AI, College of Artificial Intelligence, Gwangju Institute of Science and Technology, 123 Cheomdangwagi-ro, Buk-gu, Gwangju 61005, Repub…. Hauptthemen laut MEDLINE: Spatial Transcriptomics, Gene Expression Profiling, Deep Learning, Image Processing, Computer-Assisted. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: National Research Foundation of Korea, Institute of Information & Communications Technology Planning & Evaluation, AI Star Fellowship Support Program.
Abstract (Originalsprache)
Spatial transcriptomics (ST) enables the spatial mapping of gene expression by combining tissue images with spot-level gene expression profiles, allowing for the analysis of gene distribution and interactions between adjacent regions. Although numerous deep learning models have been developed for ST, most existing approaches are task-specific, and multimodal foundation models with broad knowledge of ST remain scarce. While a few ST multimodal models currently exist, they have been either trained on low-resolution tissue images or developed for a specific platform, thereby limiting their generalizability when applied to diverse ST data. To address this gap, we propose ST-ConMa, a multimodal foundation model pretrained on large-scale spot-level tissue images paired with corresponding gene expression profiles using contrastive and matching learning. While conventional task-specific and foundation models generate modality-specific embeddings for images and genes separately, our model jointly learns image–gene fusion representations using a multimodal encoder equipped with cross-attention and a matching objective. Through extensive evaluations on histopathology image classification benchmarks and domain-specific tasks including gene expression prediction and spatial clustering, we demonstrate that ST-ConMa learns high-quality multimodal representations and consistently outperforms existing methods. These results indicate that ST-ConMa serves as a generalizable backbone for diverse ST approaches.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of AI, College of Artificial Intelligence, Gwangju Institute of Science and Technology, 123 Cheomdangwagi-ro, Buk-gu, Gwangju 61005, Republic of Korea.
ORCID-Profile: Hyunju Lee
Förderung
- National Research Foundation of Korea
- Institute of Information & Communications Technology Planning & Evaluation
- AI Star Fellowship Support Program
- Artificial Intelligence Graduate School Program
- Gwangju Institute of Science and Technology
- Government of the Republic of Korea (Ministry of Science and ICT)
Zitieren
APA
Yeom, T., Choi, D., & Lee, H. (2026). ST-ConMa: a multimodal foundation framework for spatial transcriptomics via image–gene contrastive and matching learning. Briefings in Bioinformatics, 27(5). https://doi.org/10.1093/bib/bbag535
Vancouver
Yeom T, Choi D, Lee H. ST-ConMa: a multimodal foundation framework for spatial transcriptomics via image–gene contrastive and matching learning. Briefings in Bioinformatics. 2026;27(5). doi:10.1093/bib/bbag535
BibTeX
@article{Yeom2026,
author = {Yeom, Taelim and Choi, Dongha and Lee, Hyunju},
title = {ST-ConMa: a multimodal foundation framework for spatial transcriptomics via image–gene contrastive and matching learning},
journal = {Briefings in Bioinformatics},
year = {2026},
volume = {27},
number = {5},
doi = {10.1093/bib/bbag535}
}
EUVIMED