{
  "abstract": "A key promise of sequence-to-functon(S2F) models is their ability to evaluate arbitrary sequence inputs, providing a robust framework for understanding genotype-phenotype relationships. However, despite strong performance across genomic loci, S2F models struggle with predicting inter-individual variation in gene expression. Training a model to make genotype-dependent predictions at a single locus—an approach we call personal genome training—offers a potential solution. We introduce SAGE-net, a scalable framework and software package for training and evaluating S2F models using personal genomes. Leveraging its scalability, we conduct extensive experiments on model and training hyperparameters, demonstrating that training on personal genomes improves predictions for held-out individuals. However, the model achieves this by identifying predictive variants rather than learning a cis-regulatory grammar that generalizes across loci. This failure to generalize persists across a range of hyperparameter settings. These findings highlight the need for further exploration to unlock the full potential of S2F models in decoding the regulatory grammar of personal genomes. Scalable software and infrastructure development will be critical to this progress.",
  "abstract_shown": true,
  "affiliations": [
    "Paul G. Allen School of Computer Science and Engineering, University of Washington, Seattle, WA, USA",
    "Department of Computational and Systems Biology, University of Pittsburgh, Pittsburgh, PA, USA",
    "Genentech, South San Francisco, CA, USA. saramos@cs.washington.edu",
    "Paul G. Allen School of Computer Science and Engineering, University of Washington, Seattle, WA, USA. saramos@cs.washington.edu"
  ],
  "article_type": null,
  "arxiv": null,
  "author_total": 7,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "Anna E. Spiro",
      "orcid": "0000-0002-3465-4606"
    },
    {
      "affs": [
        0
      ],
      "name": "Xinming Tu",
      "orcid": "0009-0004-0833-6876"
    },
    {
      "affs": [
        0
      ],
      "name": "Yilun Sheng",
      "orcid": "0009-0000-0263-2616"
    },
    {
      "affs": [
        0
      ],
      "name": "Alexander Sasse"
    },
    {
      "affs": [
        1
      ],
      "name": "Rezwan Hosseini",
      "orcid": "0000-0001-6207-4628"
    },
    {
      "affs": [
        1
      ],
      "name": "Maria D. Chikina",
      "orcid": "0000-0003-2550-5403"
    },
    {
      "affs": [
        2,
        3
      ],
      "name": "Sara Mostafavi",
      "orcid": "0000-0003-4698-1177"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/f974j8s14xpej.bib",
  "cited_by": 12,
  "doi": "10.1038/s41592-026-03124-8",
  "fulltext_url": null,
  "id": "f974j8s14xpej",
  "issns": [
    "1548-7091",
    "1548-7105"
  ],
  "landing_url": "https://doi.org/10.1038/s41592-026-03124-8",
  "language": "en",
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "journal",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "1311|Genetic Associations and Epidemiology"
    ],
    [
      "openalex",
      "1312|Genomics and Chromatin Dynamics"
    ]
  ],
  "openalex": "W7163911915",
  "pdf_url": null,
  "pmcid": "PMC13345963",
  "pmid": "42260311",
  "published": "2026-06-08",
  "publisher": "Nature Portfolio",
  "ris": "https://research.draconitic.com/paper/f974j8s14xpej.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T13:26:46Z",
      "native_id": "W7163911915",
      "source": "openalex",
      "url": "https://doi.org/10.1038/s41592-026-03124-8"
    }
  ],
  "subjects": [
    {
      "field": "biochemistry",
      "origin": "openalex:1312|Genomics and Chromatin Dynamics",
      "subfield": "molecular-biology"
    },
    {
      "field": "biology",
      "origin": "openalex:1311|Genetic Associations and Epidemiology",
      "subfield": "genetics-genomics"
    }
  ],
  "title": "A scalable approach to investigating sequence-to-function predictions from personal genomes",
  "title_text": "A scalable approach to investigating sequence-to-function predictions from personal genomes",
  "type": "peer-reviewed",
  "updated_at": "2026-11-09T13:21:07Z",
  "url": "https://research.draconitic.com/paper/f974j8s14xpej",
  "venue": "Nature Methods",
  "versions": []
}