cff-version: 1.2.0 message: "Please cite the published Nature Methods paper when using this collection." title: "scFM-dataselection recovered checkpoint collection" type: dataset version: "pretraining-release" url: "https://huggingface.co/microsoft/scFM-dataselection" license: MIT authors: - family-names: "DenAdel" given-names: "Alan" - family-names: "Hughes" given-names: "Madeline" - family-names: "Thoutam" given-names: "Akshaya" - family-names: "Gupta" given-names: "Anay" - family-names: "Navia" given-names: "Andrew W." - family-names: "Fusi" given-names: "Nicolo" - family-names: "Raghavan" given-names: "Srivatsan" - family-names: "Winter" given-names: "Peter S." - family-names: "Amini" given-names: "Ava P." - family-names: "Crawford" given-names: "Lorin" preferred-citation: type: article authors: - family-names: "DenAdel" given-names: "Alan" - family-names: "Hughes" given-names: "Madeline" - family-names: "Thoutam" given-names: "Akshaya" - family-names: "Gupta" given-names: "Anay" - family-names: "Navia" given-names: "Andrew W." - family-names: "Fusi" given-names: "Nicolo" - family-names: "Raghavan" given-names: "Srivatsan" - family-names: "Winter" given-names: "Peter S." - family-names: "Amini" given-names: "Ava P." - family-names: "Crawford" given-names: "Lorin" title: "Evaluating the role of pretraining dataset size and diversity on single-cell foundation model performance" journal: "Nature Methods" year: 2026 volume: "23" issue: "7" start: "1447" end: "1457" doi: "10.1038/s41592-026-03120-y"