Zeisel Brain Anomaly Detection Benchmark Results

A precomputed benchmark of detectAnomaly (Isolation Forest) and calculateReconstructionError on the full Zeisel mouse brain dataset, evaluating detection accuracy against known ground truth (a withheld cell type) across label-noise, class-imbalance, and batch-effect gradients, plus a hyperparameter grid search. Used in the ZeiselBenchmarking vignette to illustrate how these functions perform beyond a single worked example.

Usage

zeisel_benchmark_results

Format

An object of class list of length 3.

Source

Computed from the Zeisel mouse brain dataset (scRNAseq package) using scDiagnostics::detectAnomaly and scDiagnostics::calculateReconstructionError.

Details

A named list with three data frames:

• gradients: AUROC, AUPRC, sensitivity, and specificity for both methods across 3 baseline scenarios (a distinct, a related, and a rare cell type withheld from the reference) and, for the related and rare scenarios, across gradients of label noise (0-40 reference class imbalance (300 to 10 cells), and query batch effects (a mean shift applied to 20

• if_tuning: sensitivity/specificity of detectAnomaly across a grid of PC subsets, HVG counts, and anomaly thresholds.

• re_tuning: sensitivity/specificity of calculateReconstructionError across a grid of HVG counts, PC subsets, and MAD thresholds.

Computed once, offline, on the full (non-downsampled) Zeisel dataset; see inst/script/ZeiselBenchmarkResults.R for the exact procedure.

References

Zeisel A, et al. (2015). Cell types in the mouse cortex and hippocampus revealed by single-cell RNA-seq. Science 347(6226):1138-42.

See also

Use data("zeisel_benchmark_results") to load and access the benchmark results.

Examples

# Load and explore the Zeisel benchmark results
data("zeisel_benchmark_results")