ST-ConMa: a multimodal foundation framework for spatial transcriptomics via image–gene contrastive and matching learning
Taelim Yeom, Dongha Choi, Hyunju Lee · 2026
Read ST-ConMa: a multimodal foundation framework for spatial transcriptomics via image–gene contrastive and matching learning by Taelim Yeom, Dongha Choi, Hyunju Lee (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13617578.txt.
Description
Taelim Yeom, Dongha Choi, and Hyunju Lee build ST-ConMa, a foundation model that learns tissue images and gene expression as one tongue. Spatial transcriptomics pins gene activity to spots on a tissue slice, joining picture and profile. Most models serve one task, and the few broad ones train on blurry images or a single platform.
Their 2026 framework pre-trains on large sets of spot images paired with their gene profiles. Contrastive and matching losses teach the model when a picture and a profile belong together. A joint encoder with cross-attention fuses the two streams, unlike rivals that embed each apart. Tests span image sorting, gene level guesses, and spatial grouping.
Catalog record
| CREATOR | Taelim Yeom, Dongha Choi, Hyunju Lee |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |