Article

Cross-Tissue Generalization of Rare Cell-Type Representations in Single-Cell Foundation Models: A Benchmark Using Public Reference Atlases

2026-04-25

Abstract excerpt

Single-cell foundation models (scFMs) pretrained on massive, diverse transcriptomic atlases have demonstrated remarkable performance on downstream tasks such as cell-type annotation and gene regulatory inference. However, their ability to generalize to rare cell types — populations that constitute less than 1% of cells in a given tissue — remains poorly characterized, particularly when those rare types are queried...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: af386a92-7dd4-5171-a6c0-2d18a5cd45c4
DOI: 10.14293/pr2199.003456.v1

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Cross-Tissue Generalization of Rare Cell-Type Representations in Single-Cell Foundation Models: A Benchmark Using Public Reference AtlasesDOI 10.14293/pr2199.003456.v1Read

UniCell: Towards a Unified Solution for Cell Annotation, Nomenclature Harmonization, Atlas Construction in Single-Cell TranscriptomicsSemantic proximity Read

Universal Cell Embeddings: A Foundation Model for Cell BiologySemantic proximity Read

Species-specific small models for cell type classification approach the performance of large single cell foundation modelsSemantic proximity Read

Learning interpretable cellular and gene signature embeddings from single-cell transcriptomic dataSemantic proximity Read

Benchmarking single-cell foundation models in a zero-shot settingSemantic proximity Read

Self-distillation contrastive learning enables clustering-free signature extraction and mapping to multimodal single-cell atlas of multimillion scaleSemantic proximity Read

Scaling Dense Representations for Single Cell with Transcriptome-Scale ContextSemantic proximity Read

Scaling cross-tissue single-cell annotation modelsSemantic proximity Read

Prototype-based continual cell-type annotation reveals cellular state transitions in expanding single-cell atlasesSemantic proximity Read

A Cross-Species Generative Cell Atlas Across 1.5 Billion Years of Evolution: The TranscriptFormer Single-cell ModelSemantic proximity Read

Stack: In-Context Learning of Single-Cell BiologySemantic proximity Read

Construction of continuously expandable single-cell atlases through integration of heterogeneous datasets in a generalized cell-embedding spaceSemantic proximity Read