Subcellularly Resolved Single-Cell Embedding Learning with Transcriptomic data, Protein Structure and Localization Information
Abstract
Existing cell embedding methods predominantly rely on transcriptomic or proteomic measurements and represent each cell as a holistic entity, thereby overlooking the subcellular localization of individual molecules. Moreover, they rarely incorporate protein structural information, despite its fundamental role in determining molecular interactions and functions. In this work, we propose a multimodal framework for learning subcellularly resolved cell embeddings by jointly leveraging RNA expression profiles, protein sequence representations, and protein structural information. Specifically, we employ a cross-attention architecture to integrate transcriptomic, sequence, and structural modalities and model their interactions within distinct subcellular compartments. The resulting embeddings represent each cell through its fine-grained subcellular organization, capturing both molecular expression patterns and the functional properties of the associated proteins. By learning cell representations at subcellular resolution, our framework preserves spatially organized biological information while integrating complementary signals across multiple molecular levels. To the best of our knowledge, this is the first framework that produces subcellularly resolved cell embeddings by jointly incorporating transcriptomic information, protein sequence representations, and protein structural knowledge within a unified cross-modal learning paradigm.
Related
- Cerebrovascular Structure-Flow Gradients Revealed by Volumetric Ultrasound Localization Microscopy
- Slide-level batch structure limits histology-guided supervision of transcriptomic foundation models
- FORGE audits residue-level information encoded in RNA tertiary structure geometry
- Understanding Gene Expression-to-Image Learning: Spatial Structure and Information Fidelity
- Accounting for spatial structure and surface conditions through in situ soil spectral information: application to soil organic carbon modelling
BibTeX
@misc{zhou2026subcellularly,
title = {{Subcellularly Resolved Single-Cell Embedding Learning with Transcriptomic data, Protein Structure and Localization Information}},
author = {Zhen Zhou and Jiachen Li and Yuan Liu and Xiaoyong Pan and Hong-Bin Shen},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.02344},
eprint = {2609.02344},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.02344},
}
RIS
TY - UNPB
TI - Subcellularly Resolved Single-Cell Embedding Learning with Transcriptomic data, Protein Structure and Localization Information
AU - Zhen Zhou
AU - Jiachen Li
AU - Yuan Liu
AU - Xiaoyong Pan
AU - Hong-Bin Shen
PY - 2026
DA - 2026/09/02
PB - arXiv
DO - 10.48550/arxiv.2609.02344
UR - https://arxiv.org/abs/2609.02344
AB - Existing cell embedding methods predominantly rely on transcriptomic or proteomic measurements and represent each cell as a holistic entity, thereby overlooking the subcellular localization of individual molecules. Moreover, they rarely incorporate protein structural information, despite its fundamental role in determining molecular interactions and functions. In this work, we propose a multimodal framework for learning subcellularly resolved cell embeddings by jointly leveraging RNA expression profiles, protein sequence representations, and protein structural information. Specifically, we employ a cross-attention architecture to integrate transcriptomic, sequence, and structural modalities and model their interactions within distinct subcellular compartments. The resulting embeddings represent each cell through its fine-grained subcellular organization, capturing both molecular expression patterns and the functional properties of the associated proteins. By learning cell representations at subcellular resolution, our framework preserves spatially organized biological information while integrating complementary signals across multiple molecular levels. To the best of our knowledge, this is the first framework that produces subcellularly resolved cell embeddings by jointly incorporating transcriptomic information, protein sequence representations, and protein structural knowledge within a unified cross-modal learning paradigm.
ER -