{
  "abstract": "Spatial transcriptomics pairs spatially resolved gene expression with tissue morphology in the same tissue section. Transcriptomic foundation models encode such expression profiles into general-purpose representations, but these representations can retain slide- and cohort-specific variation that obscures biological signal. Here, we test whether matched H\\&E histology can improve these representations as a training-time supervisory signal that is discarded at inference. Across three transcriptomic foundation models, histology-guided supervision provides no consistent aggregate improvement in cross-donor annotation transfer, despite substantially stronger transfer from histology alone. We find that gene-expression embeddings from spot-based spatial transcriptomics are low-dimensional and strongly structured by slide identity, limiting the shared gene--morphology signal available for cross-modal transfer. Guidance benefits some morphologically distinctive classes but reduces held-out gene predictivity broadly across the transcriptome. These results suggest that cross-modal supervision can reorganize information already encoded in a frozen representation but cannot recover information the representation does not retain, highlighting the importance of diagnosing slide-specific structure before applying such supervision. Code availability: https://github.com/ratschlab/vision-guided-transcriptomics-fm-2026",
  "abstract_shown": true,
  "affiliations": [
    "ETH Zurich"
  ],
  "article_type": "New results",
  "arxiv": null,
  "author_total": 4,
  "authors": [
    {
      "name": "U. McConnell"
    },
    {
      "affs": [
        0
      ],
      "name": "K. Nonchev"
    },
    {
      "name": "V. H. Koelzer"
    },
    {
      "name": "G. Raetsch"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/67yms4ex8fkh5.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.09.30.755585",
  "fulltext_url": "https://www.biorxiv.org/content/10.64898/2026.09.30.755585v1.full",
  "id": "67yms4ex8fkh5",
  "issns": [],
  "landing_url": "https://www.biorxiv.org/content/10.64898/2026.09.30.755585v1",
  "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": "W7220380997",
  "pdf_url": "https://www.biorxiv.org/content/10.64898/2026.09.30.755585v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-30",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/67yms4ex8fkh5.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:26:05Z",
      "native_id": "10.64898/2026.09.30.755585",
      "source": "biorxiv",
      "url": "https://www.biorxiv.org/content/10.64898/2026.09.30.755585v1"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "biorxiv:bioinformatics",
      "subfield": "bioinformatics"
    }
  ],
  "title": "Slide-level batch structure limits histology-guided supervision of transcriptomic foundation models",
  "title_text": "Slide-level batch structure limits histology-guided supervision of transcriptomic foundation models",
  "type": "preprint",
  "updated_at": "2026-10-30T12:26:05Z",
  "url": "https://research.draconitic.com/paper/67yms4ex8fkh5",
  "venue": "bioRxiv",
  "versions": []
}