Benchmark datasets for Spotless

In standards.tar.gz, the silver standards are formatted as silver_standard_i-j, where i refers to the originating single-cell dataset and j refers to the synthspot abundance pattern. Each directory of the silver standard contains 10 replicates. All reference datasets are in the 'reference/' folder.

Single-cell dataset:

Brain cortex

Single-cell cerebellum

Single-nucleus cerebellum

Hippocampus

Kidney

Squamous cell carcinoma

Melanoma

Abundance patterns:

Uniform distinct

Diverse distinct

Uniform overlap

Diverse overlap

Dominant cell type

Partially dominant cell type

Rare cell type

Regionally rare cell type

Missing cell types

We only uploaded .rds files to keep the file size smaller. For more information on file conversion, please visit the GitHub repository.

To use liver_dataset.tar.gz, please read liver_README.txt before proceeding.

dataset_visualizations.zip contain visualizations of the gold standard, silver standard, and case study datasets.

raw_results.zip contains the proportions and evaluation metrics of each method, as well as aggregated metrics used to generate plots. They can be used in tandem with the scripts in the GitHub repository.