example_config.toml

# Example TOML configuration file for xenium_process pipeline
#
# Each command has its own section. CLI arguments override config values.
# Use this file to make your pipelines reproducible across runs.
#
# Usage:
#   xenium_process concat --config example_config.toml --input samples.csv
#   xenium_process normalize --config example_config.toml --input merged.zarr
#   etc.

[concat]
# Path to CSV file listing samples (columns: sample, path, [metadata...])
input = "samples.csv"
# Path to output concatenated .zarr file
output = "merged.zarr"
# Fraction of cells to keep (0-1, default: 1.0 = no downsampling)
downsample = 1.0

[normalize]
# Path to input .zarr file
input = "merged.zarr"
# Modify the input file in place instead of creating a new file
inplace = false
# Path to output .zarr file (required unless inplace = true)
output = "normalized.zarr"
# Minimum number of genes expressed per cell
min_genes = 100
# Minimum number of cells expressing a gene
min_cells = 3
# Number of highly variable genes to select
n_top_genes = 2000
# Generate and save QC plots
save_plots = false

[cluster]
# Path to input normalized .zarr file
input = "normalized.zarr"
# Modify the input file in place instead of creating a new file
inplace = false
# Path to output .zarr file (required unless inplace = true)
output = "clustered.zarr"
# Leiden clustering resolution(s), comma-separated for multiple resolutions
leiden_resolution = "0.2,0.5,1.0"
# Generate and save UMAP plots
save_plots = true

[annotate]
# Path to input clustered .zarr file
input = "clustered.zarr"
# Modify the input file in place instead of creating a new file
inplace = false
# Path to output .zarr file (required unless inplace = true)
output = "annotated.zarr"
# Path to CSV file with marker genes (columns: cell_type, gene)
markers = "markers.csv"
# Cluster column key to use for annotation (e.g., "leiden_res0p5")
# If null, will use all leiden_res* columns found
cluster_key = null
# Pre-calculate MLM enrichment scores for pathway/TF resources
calculate_ulm = true
# Minimum sensitivity for PanglaoDB markers in MLM (default: 0.5)
panglao_min_sensitivity = 0.5
# Minimum number of marker genes per cell type for MLM annotation
tmin = 2
# Generate and save annotation plots
save_plots = true

[differential]
# Path to input .zarr file with annotations
input = "annotated.zarr"
# Directory to save differential analysis results
output_dir = "results/"
# Column in obs to group by for differential analysis
# (e.g., "status", "cell_type", or "leiden_res0p5")
groupby = "leiden_res0p5"
# Comma-separated list of exactly 2 groups to compare (Mode A)
# E.g., "HIV,NEG". If null, finds markers for all groups (Mode B)
compare_groups = null
# Optional obsm layer to use for enrichment-based differential analysis
# (e.g., "score_mlm_PanglaoDB")
obsm_layer = null
# Statistical test method for gene expression DE
# Options: "wilcoxon", "t-test", "logreg"
method = "wilcoxon"
# Layer to use for gene expression (null uses .X)
layer = null
# Generate and save differential analysis plots
save_plots = false
# Number of top genes to save per group
n_genes = 100