{"scope":{"layer":"general","tissue_scope":"unknown","species":"unknown","system":"unknown","evidence_level":"preprint","importance":"","importance_reason":""},"id":"VC0169","type":"preprint","title":"Benchmarking single-cell foundation models reveals an advantage in the low-data regime for cell-type and disease-state prediction","authors_or_org":"Benhayon H, Danziger MM, Ben-Zvi D, Rosen-Zvi M.","date":"2026-10-12","year":"2026","venue":"bioRxiv (preprint, not peer-reviewed)","url":"https://doi.org/10.64898/2026.10.12.758421","doi":"10.64898/2026.10.12.758421","pmid":"","brain_tumor_focus":"no","tags":["benchmark","foundation-model","scrna-seq","zero-shot"],"tag_slugs":["benchmark","foundation-model","scrna-seq","zero-shot"],"tag_labels":["Benchmark & validation","Foundation model","scRNA-seq","Zero-shot"],"tag_labels_zh":["基准评测与验证","基础模型","单细胞 RNA 测序","零样本"],"unmapped_tags":[],"one_line_summary_en":"Preprint (not peer-reviewed): Benhayon et al. (Institute for Medical Research Israel-Canada, Jerusalem) benchmark cell representations from four single-cell foundation models with logistic regression on 868 disease-prediction tasks (82 datasets) and 65 cell-type datasets from CELLxGENE. With full data, the log-normalized-count baseline outperformed the best model (TranscriptFormer-Sapiens) for cell type and matched it for disease; with few labelled cells, TranscriptFormer-Sapiens significantly outperformed the baseline in both tasks up to about 10^2 samples per class.","one_line_summary_zh":"预印本（未经同行评审）：Benhayon 等（以色列-加拿大医学研究所，耶路撒冷）在 CELLxGENE 的 868 个疾病预测任务（82 个数据集）和 65 个细胞类型数据集上，用逻辑回归评测 4 个单细胞基础模型的细胞表征。全量数据时，log 归一化计数基线在细胞类型上优于最佳模型（TranscriptFormer-Sapiens），在疾病预测上与之持平；标注细胞较少时（每类约 10² 个以内），TranscriptFormer-Sapiens 在两类任务上均显著优于基线。","verified":"yes","verification_source":"bioRxiv API details (title, 4 authors, posted 2026-10-12, v1, new results, CC BY, corresponding institution IMRIC Jerusalem, abstract) fetched 2026-10-20; full text not read; KB id VC0169 (renumbered 2026-10-20)","page_url":"https://gliostate.com/library/VC0169/","api_url":"https://gliostate.com/api/items/VC0169.json","sameAs":["https://doi.org/10.64898/2026.10.12.758421"],"is_preprint":true,"facets":{"type":"https://gliostate.com/library/type/preprint/","year":"https://gliostate.com/library/year/2026/","tags":["https://gliostate.com/library/tag/benchmark/","https://gliostate.com/library/tag/foundation-model/","https://gliostate.com/library/tag/scrna-seq/","https://gliostate.com/library/tag/zero-shot/"]},"cited_by":[],"related":[{"id":"VC0166","title":"PIE: Generalizing perturbation effects across unseen perturbations, contexts and datasets","shared_tags":["benchmark","zero-shot"],"page_url":"https://gliostate.com/library/VC0166/","api_url":"https://gliostate.com/api/items/VC0166.json"},{"id":"VC0143","title":"Generating high-quality Perturb-seq data for the 2026 Virtual Cell Challenge","shared_tags":["benchmark","zero-shot"],"page_url":"https://gliostate.com/library/VC0143/","api_url":"https://gliostate.com/api/items/VC0143.json"},{"id":"VC0058","title":"Virtual Cell Challenge 2026: Benchmarking zero-shot generalization across cellular contexts.","shared_tags":["benchmark","zero-shot"],"page_url":"https://gliostate.com/library/VC0058/","api_url":"https://gliostate.com/api/items/VC0058.json"},{"id":"VC0050","title":"Zero-shot evaluation reveals limitations of single-cell foundation models.","shared_tags":["benchmark","zero-shot"],"page_url":"https://gliostate.com/library/VC0050/","api_url":"https://gliostate.com/api/items/VC0050.json"},{"id":"VC0107","title":"Virtual Cell Challenge 2026 (zero-shot, 6 unseen cell lines)","shared_tags":["zero-shot"],"page_url":"https://gliostate.com/library/VC0107/","api_url":"https://gliostate.com/api/items/VC0107.json"},{"id":"VC0174","title":"Emerging frontiers for AI in single-cell genomics","shared_tags":["foundation-model","scrna-seq"],"page_url":"https://gliostate.com/library/VC0174/","api_url":"https://gliostate.com/api/items/VC0174.json"}]}