{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T21:19:30Z","timestamp":1792790370923,"version":"4.5.0"},"posted":{"date-parts":[[2026,10,1]]},"group-title":"Bioinformatics","reference-count":78,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.biorxiv.org\/about\/FAQ#license"}],"funder":[{"name":"National Institute of General Medical Sciences of the National Institutes of Health","award":["R35GM151089"],"award-info":[{"award-number":["R35GM151089"]}],"id":[{"id":"https:\/\/ror.org\/04q48ey07","id-type":"ROR","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,10,1]]},"abstract":"<jats:title>\n                  A\n                  <jats:sc>bstract<\/jats:sc>\n                <\/jats:title>\n                <jats:p>\n                  Spatial gene expression prediction from histology is typically evaluated by the mean per-gene Pearson correlation (PCC). PCC does not distinguish spatial scales: coarse tissue organisation alone can earn most of the score. A domain oracle makes this precise: given an image-derived partition of the tissue, the predictor that assigns each domain its measured mean expression is the best that any predictor constant on that partition can do. On a widely used benchmark with a \u223c100\n                  <jats:italic>\u00b5<\/jats:italic>\n                  m pitch, a learnable version that estimates each domain\u2019s mean from training data reaches 82% of the PCC of a trained model, and a partition of the spot coordinates alone, given measured means, reaches 123%. At that pitch the score rewards coarse spatial structure whatever its source. We introduce O\n                  <jats:sc>ctave<\/jats:sc>\n                  , a scale-resolved evaluation that filters prediction and measurement into spatial bands with widths in micrometres and scores their agreement in each band. On 16\n                  <jats:italic>\u00b5<\/jats:italic>\n                  m Xenium data, differences that look modest under PCC grow at finer scales: at \u223c20\n                  <jats:italic>\u00b5<\/jats:italic>\n                  m the oracle\u2019s shortfall relative to the trained model is 2.4 times its PCC shortfall at the median and exceeds it in all 13 specimens. Across 57 image encoders the finest band reorders 261 of the 1596 encoder pairs relative to PCC. The benchmark\u2019s own finest band is 50\n                  <jats:italic>\u00b5<\/jats:italic>\n                  m wide, coarser than the scale at which these differences appear, so no score on that pitch can directly validate finer structure. O\n                  <jats:sc>ctave<\/jats:sc>\n                  reports how well a model predicts spatial expression at each spatial scale.\n                <\/jats:p>","DOI":"10.64898\/2026.09.25.754484","type":"posted-content","created":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T00:40:26Z","timestamp":1790901626000},"source":"Crossref","is-referenced-by-count":0,"title":["Octave: Scale-Resolved Evaluation of Spatial Gene Expression Prediction from Histology"],"prefix":"10.64898","author":[{"given":"Qing","family":"Wang","sequence":"first","affiliation":[{"name":"Department of Health Outcomes and Biomedical Informatics, University of Florida, Gainesville, FL, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sipei","family":"Gu","sequence":"additional","affiliation":[{"name":"Department of Health Outcomes and Biomedical Informatics, University of Florida, Gainesville, FL, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qizhen","family":"Lan","sequence":"additional","affiliation":[{"name":"Department of Health Data Science and Artificial Intelligence, McWilliams School of Biomedical Informatics, UTHealth Houston, Houston, TX, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaoqian","family":"Jiang","sequence":"additional","affiliation":[{"name":"Department of Health Data Science and Artificial Intelligence, McWilliams School of Biomedical Informatics, UTHealth Houston, Houston, TX, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yijiang","family":"Chen","sequence":"additional","affiliation":[{"name":"Department of Medicine, College of Medicine, University of Florida, Gainesville, FL, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4455-5302","authenticated-orcid":false,"given":"Qianqian","family":"Song","sequence":"additional","affiliation":[{"name":"School of Applied and Creative Computing, Purdue University, West Lafayette, IN, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026100610051158000_2026.09.25.754484v1.1","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-025-02772-6"},{"key":"2026100610051158000_2026.09.25.754484v1.2","article-title":"A deployment-friendly foundational framework for efficient computational pathology","year":"2026","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.3","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-024-02857-3"},{"key":"2026100610051158000_2026.09.25.754484v1.4","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-025-02707-1"},{"key":"2026100610051158000_2026.09.25.754484v1.5","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.01101"},{"key":"2026100610051158000_2026.09.25.754484v1.6","doi-asserted-by":"publisher","DOI":"10.1016\/j.mlwa.2021.100198"},{"key":"2026100610051158000_2026.09.25.754484v1.7","doi-asserted-by":"publisher","DOI":"10.1016\/j.acha.2006.04.006"},{"key":"2026100610051158000_2026.09.25.754484v1.8","doi-asserted-by":"publisher","DOI":"10.1016\/j.acha.2006.04.004"},{"key":"2026100610051158000_2026.09.25.754484v1.9","article-title":"Vision transformers need registers","year":"2024","journal-title":"In International Conference on Learning Representations"},{"key":"2026100610051158000_2026.09.25.754484v1.10","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-025-03982-3"},{"key":"2026100610051158000_2026.09.25.754484v1.11","doi-asserted-by":"publisher","DOI":"10.1101\/2023.07.21.23292757"},{"key":"2026100610051158000_2026.09.25.754484v1.12","article-title":"Phikon-v2, a large and public feature extractor for biomarker prediction","year":"2024","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.13","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-032-04981-016"},{"key":"2026100610051158000_2026.09.25.754484v1.14","article-title":"Robustifying pathology foundation models via fine-tuning","year":"2026","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.15","doi-asserted-by":"publisher","DOI":"10.1109\/iccvw69036.2025.00128"},{"key":"2026100610051158000_2026.09.25.754484v1.16","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-032-37029-76"},{"key":"2026100610051158000_2026.09.25.754484v1.17","doi-asserted-by":"publisher","DOI":"10.1016\/j.acha.2010.04.005"},{"key":"2026100610051158000_2026.09.25.754484v1.18","doi-asserted-by":"publisher","DOI":"10.1038\/s41551-020-0578-x"},{"key":"2026100610051158000_2026.09.25.754484v1.19","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-023-02504-3"},{"key":"2026100610051158000_2026.09.25.754484v1.20","doi-asserted-by":"publisher","DOI":"10.2307\/1942661"},{"key":"2026100610051158000_2026.09.25.754484v1.21","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1654"},{"key":"2026100610051158000_2026.09.25.754484v1.22","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-023-43458-x"},{"key":"2026100610051158000_2026.09.25.754484v1.23","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1704"},{"key":"2026100610051158000_2026.09.25.754484v1.24","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbae464"},{"key":"2026100610051158000_2026.09.25.754484v1.25","article-title":"Towards large-scale training of pathology foundation models","author":"kaiko.ai","year":"2024","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.26","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52729.2023.00326"},{"key":"2026100610051158000_2026.09.25.754484v1.27","unstructured":"Daniel Kaplan , Ratna Sagari Grandhi , Connor Lane , Benjamin Warner , Tanishq Mathew Abraham , and Paul S. Scotti . OpenMidnight: How to train a state-of-the-art pathology foundation model with $1.6k. Sophont technical report, https:\/\/sophont.med\/blog\/openmidnight, 2025. Model release, https:\/\/huggingface.co\/SophontAI\/OpenMidnight."},{"key":"2026100610051158000_2026.09.25.754484v1.28","doi-asserted-by":"publisher","DOI":"10.64898\/2026.03.17.712534"},{"key":"2026100610051158000_2026.09.25.754484v1.29","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-032-04984-155"},{"key":"2026100610051158000_2026.09.25.754484v1.30","article-title":"DistillPath: An efficient 22M distilled pathology encoder approaching large foundation model performance","author":"Accepted at the ECCV 2026 Workshop on Medical Foundation Models and Benchmarks","year":"2026","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.31","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-026-73923-2"},{"key":"2026100610051158000_2026.09.25.754484v1.32","article-title":"Domain-specific optimization and diverse evaluation of self-supervised models for histopathology","year":"2023","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.33","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbae500"},{"key":"2026100610051158000_2026.09.25.754484v1.34","doi-asserted-by":"publisher","DOI":"10.1002\/advs.202514351"},{"key":"2026100610051158000_2026.09.25.754484v1.35","doi-asserted-by":"publisher","DOI":"10.1080\/757582976"},{"key":"2026100610051158000_2026.09.25.754484v1.36","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-024-02856-4"},{"key":"2026100610051158000_2026.09.25.754484v1.37","doi-asserted-by":"publisher","DOI":"10.1038\/s41551-025-01488-4"},{"key":"2026100610051158000_2026.09.25.754484v1.38","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72083-39"},{"key":"2026100610051158000_2026.09.25.754484v1.39","doi-asserted-by":"publisher","DOI":"10.1093\/biomet\/37.1-2.17"},{"key":"2026100610051158000_2026.09.25.754484v1.40","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58595-241"},{"key":"2026100610051158000_2026.09.25.754484v1.41","article-title":"Hibou: A family of foundational vision transformers for pathology","year":"2024","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.42","doi-asserted-by":"publisher","DOI":"10.64898\/2026.05.07.723486"},{"key":"2026100610051158000_2026.09.25.754484v1.43","doi-asserted-by":"publisher","DOI":"10.1101\/2025.02.09.25321567"},{"key":"2026100610051158000_2026.09.25.754484v1.44","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-025-02193-3"},{"key":"2026100610051158000_2026.09.25.754484v1.45","article-title":"DINOv2: Learning robust visual features without supervision","year":"2024","journal-title":"Transactions on Machine Learning Research"},{"key":"2026100610051158000_2026.09.25.754484v1.46","article-title":"Disentangled shared representations improve morpho-transcriptomic integration","author":"To appear at the MICCAI 2026 COMPAYL Workshop","year":"2026","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.47","doi-asserted-by":"publisher","DOI":"10.1101\/2021.11.28.470212"},{"key":"2026100610051158000_2026.09.25.754484v1.48","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-54182-5"},{"key":"2026100610051158000_2026.09.25.754484v1.49","first-page":"8748","volume-title":"In Proceedings of the 38th International Conference on Machine Learning, volume 139 of Proceedings of Machine Learning Research","year":"2021"},{"key":"2026100610051158000_2026.09.25.754484v1.50","first-page":"5389","volume-title":"In Proceedings of the 36th International Conference on Machine Learning, volume 97 of Proceedings of Machine Learning Research","year":"2019"},{"key":"2026100610051158000_2026.09.25.754484v1.51","unstructured":"Charlie Saillard , Rodolphe Jenatton , Felipe Llinares-L\u00f3pez, Zelda Mariet, David Cahan\u00e9, Eric Durand, and Jean-Philippe Vert. H-optimus-0, 2024. URL https:\/\/github.com\/bioptimus\/releases\/tree\/main\/models\/h-optimus\/v0."},{"key":"2026100610051158000_2026.09.25.754484v1.52","doi-asserted-by":"publisher","DOI":"10.1158\/1538-7445.AM2026-LB174"},{"key":"2026100610051158000_2026.09.25.754484v1.53","doi-asserted-by":"publisher","DOI":"10.1109\/bibm62325.2024.10822048"},{"key":"2026100610051158000_2026.09.25.754484v1.54","doi-asserted-by":"publisher","DOI":"10.1109\/msp.2012.2235192"},{"key":"2026100610051158000_2026.09.25.754484v1.55","article-title":"DINOv3","year":"2026","journal-title":"Transactions on Machine Learning Research","ISSN":"https:\/\/id.crossref.org\/issn\/2835-8856","issn-type":"print"},{"key":"2026100610051158000_2026.09.25.754484v1.56","article-title":"PathGen-1.6M: 1.6 million pathology image-text pairs generation through multi-agent collaboration","year":"2025","journal-title":"In International Conference on Learning Representations"},{"key":"2026100610051158000_2026.09.25.754484v1.57","article-title":"A note on the evaluation of generative models","year":"2016","journal-title":"In International Conference on Learning Representations"},{"key":"2026100610051158000_2026.09.25.754484v1.58","article-title":"SigLIP 2: Multilingual vision-language encoders with improved semantic understanding, localization, and dense features","year":"2025","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.59","article-title":"GigaPath-Flash and GigaTIME-Flash: Efficient pathology foundation models for whole-slide and tumor microenvironment analysis","year":"2026","journal-title":"arXiv preprint"},{"key":"2026100610051158000_2026.09.25.754484v1.60","doi-asserted-by":"publisher","DOI":"10.1038\/s41587-025-02777-8"},{"key":"2026100610051158000_2026.09.25.754484v1.61","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-024-03141-0"},{"key":"2026100610051158000_2026.09.25.754484v1.62","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i7.32830"},{"key":"2026100610051158000_2026.09.25.754484v1.63","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btaf363"},{"key":"2026100610051158000_2026.09.25.754484v1.64","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2022.102559"},{"key":"2026100610051158000_2026.09.25.754484v1.65","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2022.102645"},{"key":"2026100610051158000_2026.09.25.754484v1.66","doi-asserted-by":"publisher","DOI":"10.1109\/acssc.2003.1292216"},{"key":"2026100610051158000_2026.09.25.754484v1.67","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-024-08378-w"},{"key":"2026100610051158000_2026.09.25.754484v1.68","doi-asserted-by":"publisher","DOI":"10.1093\/bioadv\/vbag202"},{"key":"2026100610051158000_2026.09.25.754484v1.69","doi-asserted-by":"publisher","DOI":"10.52202\/075280-3095"},{"key":"2026100610051158000_2026.09.25.754484v1.70","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-024-07441-w"},{"key":"2026100610051158000_2026.09.25.754484v1.71","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-66220-x"},{"key":"2026100610051158000_2026.09.25.754484v1.72","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-024-02325-3"},{"key":"2026100610051158000_2026.09.25.754484v1.73","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbac297"},{"key":"2026100610051158000_2026.09.25.754484v1.74","doi-asserted-by":"publisher","DOI":"10.1038\/s41587-023-02019-9"},{"key":"2026100610051158000_2026.09.25.754484v1.75","doi-asserted-by":"publisher","DOI":"10.1056\/AIoa2400640"},{"key":"2026100610051158000_2026.09.25.754484v1.76","doi-asserted-by":"publisher","DOI":"10.1016\/j.ccell.2026.01.019"},{"key":"2026100610051158000_2026.09.25.754484v1.77","article-title":"Diffusion generative modeling for spatially resolved gene expression inference from histology images","year":"2025","journal-title":"In International Conference on Learning Representations"},{"key":"2026100610051158000_2026.09.25.754484v1.78","article-title":"Virchow2: Scaling self-supervised mixed magnification models in pathology","year":"2024","journal-title":"arXiv preprint"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.09.25.754484","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T17:05:22Z","timestamp":1791306322000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.09.25.754484"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,1]]},"references-count":78,"URL":"https:\/\/doi.org\/10.64898\/2026.09.25.754484","relation":{},"subject":[],"published":{"date-parts":[[2026,10,1]]},"subtype":"preprint"}}