v1_0.md

Overview

We expect a marker_detection HDF5 group at the root of the file, containing marker statistics for each cluster. The group itself contains the parameters and results subgroups.

Definitions:

• num_clusters: number of clusters in the analysis. This is typically determined from choose_clustering.
• num_genes: number of genes in the analysis. This is typically determined from the inputs step.

Parameters

parameters should be empty.

Results

results should contain:

• clusters: a group representing an array of length equal to the number of clusters. Each child is another group that is named by the cluster index from 0 to num_clusters - 1, containing the marker details for that cluster. Each child group contains:
  • means: a float dataset of length equal to num_genes, containing the mean expression of each gene in the current cluster.
  • detected: a float dataset of length equal to num_genes, containing the proportion of cells with detected expression of each gene in the current cluster.
  • lfc: an group containing statistics for the log-fold changes from all pairwise comparisons involving the current cluster. This contains:
    • min: a float dataset of length equal to num_genes, containing the minimum log-fold change across all pairwise comparisons for each gene.
    • mean: a float dataset of length equal to num_genes, containing the mean log-fold change across all pairwise comparisons for each gene.
    • min_rank: a float dataset of length equal to num_genes, containing the minimum rank of the log-fold changes across all pairwise comparisons for each gene.
  • delta_detected: same as lfc, but for the delta-detected (i.e., difference in the percentage of detected expression).
  • cohen: same as lfc, but for Cohen's d.
  • auc: same as lfc, but for the AUCs.

History

Added in version 1.0.