A benchmark dataset for single-cell shape analysis with four configurations: real-world scRNA-seq data, synthetic data, annotator labels, and aggregated labels.
scShapeBench is a curated collection of datasets assembled for benchmarking computational methods in single-cell shape analysis. It is organized into four configurations:
Real-world single-cell gene expression datasets. Each dataset is stored as an individual… See the full description on the dataset page:
https://huggingface.co/datasets/scShape-Benchmark/scShapeBench.
