Overview

Detailed model evaluation on curated benchmark datasets are stored here to inform users about mapping algorithm biases and accuracies at mapping cell type labels using Allen Institute taxonomies as references.

Evaluations: Hierarchical Correlation Mapping (HANN)

Model Benchmark Runtime
HANN mouse WB 0.76 Hours
HANN human MTG SEA-AD 3 Hours

Evaluations: Correlation Mapping (CORR)

Model Benchmark Runtime
CORR human MTG SEA-AD 1.8 Hours
CORR mouse WB 0.36 Hours