Samples table
Format
A comma-separated .csv file passed with --samples, with a header row and these columns:
| Column | Required | Rules |
|---|---|---|
id | Yes | Sample name, used for the output folder. Unique across the table, no spaces. |
path | Yes | No spaces. A CellRanger or STARsolo output folder, a .h5 file or a .mtx folder; with --on_irods, an iRODS collection. |
id,path
cellranger-count,/lustre/scratch124/.../cellranger-count
cellranger-count-h5,/lustre/scratch124/.../outs/raw_feature_bc_matrix.h5
The table is checked against assets/schema_samples.json before the run starts. A duplicate id or a missing column stops it with, for example:
* --samples (samples.csv): Validation of file failed:
-> Entry 2: Detected duplicate entries: [id:s1]
The folder layouts the pipeline recognises are listed in Cell and droplet counts.
Example tables
| File | Contents |
|---|---|
examples/samples.csv | Local CellRanger and STARsolo folders, .h5 files and .mtx folders |
examples/samples_irods.csv | iRODS collections, for --on_irods |
examples/samples_preset.csv | Whole mapper folders only, for the mapper preset |
examples/samples_exclude_features.csv | A mix of library types (ARC, CRISPR, multiplexed, Flex) on iRODS, for trying --exclude_features |
examples/run_cellbender.sh is a complete run script using one of them.
Index files
Each run writes the same format back out, so its outputs can be the input of the next run:
| File | Lists |
|---|---|
results/index/cellbender.csv | The CellBender output folder of each sample |
results/index/raw.csv | The folders downloaded from iRODS (empty without --on_irods) |
"id","path"
"sample_a","/lustre/scratch124/.../results/cellbender/sample_a"
Values are quoted; the pipeline reads quoted and unquoted tables alike. index/raw.csv is the useful one to feed back: it reruns CellBender without downloading again (see Data on iRODS).