scRM: A single-cell foundation model trained on scRNA-seq reads
To our knowledge, scRM is the first single-cell foundation model trained on sequencing reads, demonstrating that raw data pretraining is not only feasible but competitive at a fraction of the computational cost. We anticipate that scaling this framework and exploring alternative architectures will unlock biological information inaccessible to count models and extend the dataset-specific pretraining it already makes possible. The source code is available at https://github.com/altos-labs/scRM.
Authors
- Rob Patro (ORCID: https://orcid.org/0000-0001-8463-1675)
- Marcel Nassar (ORCID: https://orcid.org/0000-0002-7628-8819)
- Dongze He (ORCID: https://orcid.org/0000-0001-8259-7434)
- Jamin Liu (ORCID: https://orcid.org/0000-0001-6386-3709)
- Qi Jiang Yan (ORCID: https://orcid.org/0000-0003-4418-123X)
- Stefka Tyanova (ORCID: https://orcid.org/0000-0002-1974-5876)
- Kristof Torkenczy
- Glenn Xu
- Haiyu Wu
Institutions
- Altos Labs (US)
- University of Maryland, College Park (US)
Publication Details
- Journal
- bioRxiv (Cold Spring Harbor Laboratory)
- Published
- 2026-11-02
- DOI
- https://doi.org/10.64898/2026.10.24.761590
- Primary Topic
- Single-cell and spatial transcriptomics
- Type
- preprint