Single-cell RNA sequencing has changed how scientists explore the inner workings of cells by measuring gene activity in individual cells. This technology helps identify different cell types and states, especially in diseases like cancer or immune disorders. However, even the most advanced computational models tend to simplify cells into fixed categories that overlook their complex biological context.
Researchers from École Polytechnique, IP Paris have developed an innovative framework called Cell2Text that bridges the gap between gene expression data and human language. Cell2Text combines the analytical power of single-cell foundation models with the expressive capabilities of large language models. The result is a system that can “translate” single-cell RNA sequencing data into readable, structured text descriptions that summarize each cell’s identity, tissue source, disease relevance, and biological pathway activity.
Overview of the Cell2Text framework
The model takes single-cell RNA-seq profiles as input and processes them through a pretrained Geneformer encoder to generate contextualized gene-level embeddings. These embeddings are projected into the semantic space of the language model via a lightweight adapter module, aligning biological signals with linguistic representations. A pretrained, instruction-tuned LLM decoder then generates structured natural language descriptions that capture cellular identity, tissue of origin, disease associations, and pathway activity.
Unlike older models that rely on rigid labels, Cell2Text produces flexible, interpretable summaries that give researchers a clearer understanding of cell behavior. The framework not only improves the accuracy of cell classification but also enhances the interpretability of the results, allowing scientists to generate meaningful descriptions even for previously unseen cell types.
This integration of artificial intelligence and RNA sequencing provides a powerful new way to interpret complex biological data, paving the way for more efficient and insightful analysis of cellular diversity.
Availability – Anonymized codebase is available at https://anonymous.4open.science/r/cell2text-FDDF.
Kharouiche O, Markogiannakis A, Fei X, Chatzianastasis M, Vazirgiannis M. (2025) Cell2Text: Multimodal LLM for Generating Single-Cell Descriptions from RNA-Seq Data. arXiv 2509.24840v1. [article]
Related Posts
Single-cell RNA sequencing has changed how scientists explore the inner workings of cells by measuring gene activity in individual cells. This technology helps identify different cell types and states, especially in diseases like cancer or immune disorders. However, even the most advanced computational models tend to simplify cells into fixed categories that overlook their complex biological context.
Researchers from École Polytechnique, IP Paris have developed an innovative framework called Cell2Text that bridges the gap between gene expression data and human language. Cell2Text combines the analytical power of single-cell foundation models with the expressive capabilities of large language models. The result is a system that can “translate” single-cell RNA sequencing data into readable, structured text descriptions that summarize each cell’s identity, tissue source, disease relevance, and biological pathway activity.
Overview of the Cell2Text framework
The model takes single-cell RNA-seq profiles as input and processes them through a pretrained Geneformer encoder to generate contextualized gene-level embeddings. These embeddings are projected into the semantic space of the language model via a lightweight adapter module, aligning biological signals with linguistic representations. A pretrained, instruction-tuned LLM decoder then generates structured natural language descriptions that capture cellular identity, tissue of origin, disease associations, and pathway activity.
Unlike older models that rely on rigid labels, Cell2Text produces flexible, interpretable summaries that give researchers a clearer understanding of cell behavior. The framework not only improves the accuracy of cell classification but also enhances the interpretability of the results, allowing scientists to generate meaningful descriptions even for previously unseen cell types.
This integration of artificial intelligence and RNA sequencing provides a powerful new way to interpret complex biological data, paving the way for more efficient and insightful analysis of cellular diversity.
Availability – Anonymized codebase is available at https://anonymous.4open.science/r/cell2text-FDDF.
Kharouiche O, Markogiannakis A, Fei X, Chatzianastasis M, Vazirgiannis M. (2025) Cell2Text: Multimodal LLM for Generating Single-Cell Descriptions from RNA-Seq Data. arXiv 2509.24840v1. [article]
Related Posts
Recent Posts
Scientists identify the neuroblastoma cells that signal poor outcomes
GARAGE improves rare cell detection in single-cell RNA sequencing analysis using AI
Machine learning reveals how RNA splicing is regulated in individual cell types
Illumina releases SpliceAI2 to help advance rare disease research
Postdoctoral Fellow – Long-Read RNA Methods Development for Neurodegeneration
Novel biosensor uses DNA ‘barcodes’ to uncover disease signals in cells
RNA-Seq Products & Services
Stay Connected