{
  "abstract": "Single-cell foundation models typically represent cells as flat sequences of individual genes, requiring higher-order biological organization to be inferred implicitly from sparse expression profiles. Similar to the evolution from character-level to semantic tokenization in natural language processing,we hypothesize that cellular states can be more efficiently represented using biologically meaningful intermediate units. Here we present BioMod, a 50-million-parameter single-cell foundation model that introduces biological modules as representation units between genes and cells. BioMod constructs an overlapping vocabulary of 660 biological modules derived from curated pathways and data-driven gene programs, replacing thousands of gene-level inputs with compact module representations. Pretrained on approximately 73 million human cells using high-ratio module masking with an asymmetric decoder, BioMod achieves state-of-the-art performance across cell-type annotation, zero-shot batch integration, perturbation-response prediction and drug-response forecasting while maintaining a compact architecture that enables efficient inference on a single GPU. Module-level analyses further demonstrate that BioMod captures biologically meaningful cellular programs and preserves pathway-level organization.",
  "abstract_shown": true,
  "affiliations": [
    "Shenzhen Loop Area Instutite"
  ],
  "article_type": "New results",
  "arxiv": null,
  "author_total": 16,
  "authors": [
    {
      "name": "L. Meng"
    },
    {
      "name": "J. Zhu"
    },
    {
      "name": "B. Han"
    },
    {
      "name": "Y. Zang"
    },
    {
      "name": "W. Li"
    },
    {
      "name": "K. Zong"
    },
    {
      "name": "J. Zhong"
    },
    {
      "name": "Y. Yang"
    },
    {
      "name": "J. Liu"
    },
    {
      "name": "J. Jia"
    },
    {
      "name": "R. Sun"
    },
    {
      "name": "G. Yu"
    },
    {
      "name": "J. Han"
    },
    {
      "name": "Y. Chen"
    },
    {
      "name": "L. Zhang"
    },
    {
      "affs": [
        0
      ],
      "name": "J. Xu"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/4e385bms6tbwn.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.10.12.758490",
  "fulltext_url": "https://www.biorxiv.org/content/10.64898/2026.10.12.758490v1.full",
  "id": "4e385bms6tbwn",
  "issns": [],
  "landing_url": "https://www.biorxiv.org/content/10.64898/2026.10.12.758490v1",
  "language": null,
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "biorxiv",
      "bioinformatics"
    ],
    [
      "openalex",
      "1312|Single-cell and spatial transcriptomics"
    ]
  ],
  "openalex": "W7221628709",
  "pdf_url": "https://www.biorxiv.org/content/10.64898/2026.10.12.758490v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-12",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/4e385bms6tbwn.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:27:28Z",
      "native_id": "10.64898/2026.10.12.758490",
      "source": "biorxiv",
      "url": "https://www.biorxiv.org/content/10.64898/2026.10.12.758490v1"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "biorxiv:bioinformatics",
      "subfield": "bioinformatics"
    }
  ],
  "title": "BioMod: Biological Module Tokenization for Single-Cell Foundation Modeling",
  "title_text": "BioMod: Biological Module Tokenization for Single-Cell Foundation Modeling",
  "type": "preprint",
  "updated_at": "2026-10-30T12:27:28Z",
  "url": "https://research.draconitic.com/paper/4e385bms6tbwn",
  "venue": "bioRxiv",
  "versions": []
}