Create a sample sheet

For	Users who need stable sample IDs or non-default slide locations
Hands-on steps	Write two columns, inspect, correct duplicates
Prerequisites	Prepared primary slide files and neutral, non-PHI sample aliases
Download/storage	None; the CSV and inspection manifest are small
Writes to	A user-created CSV and a separate inspection output

Use UTF-8 CSV with sample_id,slide_path:

sample_id,slide_path
sample_001_block_a,case_001/1_L0_rgb.tif
sample_002_block_a,case_002/1_L0_rgb.tif

Paths may be relative to the input directory or absolute. Sample IDs become output directory and matrix-column identifiers. Do not use names, accession numbers, or other patient identifiers.

Validate it without inference

tumorquantai inspect /data/slides \
  --sample-sheet /data/slides/samples.csv \
  --output /data/tumorquantai-inspection

Expected success is a manifest containing exactly one row per sample sheet entry. Duplicate IDs, duplicate slide paths, missing files, unsafe paths, and unexpected companions must fail or be reported for review.

The sample sheet controls slide IDs; an optional aggregation mapping is a separate advanced concept. Pooling multiple slides into one biological sample must retain every included and excluded source slide in the audit.

Stop and clean up

Press Ctrl+C and rerun after editing the CSV. Remove only the separate inspection output; keep the reviewed sample sheet with run provenance.

Next: choose smoke, fast, or full.