biomarker-qc

Standalone QC module for biomarker TSV files.

Usage

python biomarker-qc.py <file.tsv>
python biomarker-qc.py <file.tsv> --report qc_report.txt

Checks

#	Check	Level
1	Required-field completeness — flags rows where any required field is missing or blank	ERROR
2	Name/ID consistency — detects cases where the same ID is paired with different display names across rows (covers condition_id, assessed_biomarker_entity_id, exposure_agent_id, specimen_id)	WARNING
3	Evidence-source format — verifies every evidence_source value follows DATABASE:ACCESSION format (e.g. PubMed:12345678)	ERROR
4	Duplicate row detection — flags exact duplicate rows	WARNING

Required fields

biomarker_id, biomarker, assessed_biomarker_entity, assessed_biomarker_entity_id, assessed_entity_type, best_biomarker_role, evidence_source, evidence

Output

A summary is always printed to stdout. Pass --report <path> to also write it to a file.

============================================================
QC SUMMARY  —  2 errors, 1 warnings, 0 info
============================================================
[ERROR] (row 4) [evidence_source]: evidence_source 'PubMed' does not match expected 'DATABASE:ACCESSION' format.
[ERROR] (row 9) [biomarker_id]: Required field 'biomarker_id' is empty.
[WARNING] (file) [condition]: ID 'DOID:1612' (condition_id) is paired with multiple names ...
============================================================

Exit codes

Code	Meaning
0	No errors (warnings may be present)
1	One or more ERRORs found