Speccheck
Turn a directory of bacterial QC outputs into one reviewable decision table.
speccheck sits after tools such as QUAST, CheckM2, Speciator, Sylph, Fastp,
BUSCO, ARIBA, or a workflow such as GHRU Assembly. It detects recognised output
files, applies an explicit criteria CSV, and writes compact CSV, HTML, and XLSX
reports that can be checked by a person and reproduced later.
-
:material-file-search-outline: Find the evidence
Detect QC files with
speccheck inspectbefore writing anything. -
:material-table-check: Apply transparent thresholds
Use species-specific rows where available and generic rows where not.
-
:material-chart-box-outline: Review a cohort
Generate concise CSV, wide CSV, XLSX, and interactive HTML reports.
-
:material-source-branch: Fit a pipeline
Use
collect-pipeline --layout ghruor wirecollectandsummaryinto a new workflow.
The core workflow
flowchart LR
A[Upstream QC tools] --> B[speccheck inspect]
B --> C[speccheck collect]
D[criteria.csv] --> C
C --> E[one CSV per sample]
E --> F[speccheck summary]
F --> G[report.csv]
F --> H[report.html]
F --> I[report.xlsx]
Install
The published package name is speccheck-qc; the command it installs is
speccheck.
python -m pip install speccheck-qc
speccheck --help
Use an editable source install only if you are developing Speccheck itself.
A minimal example
Collect one sample:
speccheck collect path/to/sample_qc/ \
--sample SAMPLE_001 \
--organism "Escherichia coli" \
--assembly-type short \
--output-file qc_collect/SAMPLE_001.csv
Summarise a folder of collected samples:
speccheck summary qc_collect \
--output qc_report \
--plot \
--xlsx-output qc_report/report.xlsx
Start with Quick Start if you have QC files already. If you are wiring a workflow, read Pipeline Integration. If you want to understand thresholds, read Criteria and Thresholds.
What makes Speccheck useful?
It does not replace upstream QC tools. It makes their outputs comparable by applying one criteria table and recording the threshold source, criteria checksum, Speccheck version, and input-file count in the output.
Worked example: 100 real read-backed genomes
The repository includes a compact 100-sample E. coli case study generated from GHRU Assembly outputs. It shows the intended use of Speccheck on a realistic cohort without committing raw reads or workflow work directories.
The committed example contains:
- accessions and selection metadata;
- Speccheck
report.csv,report.full.csv,report.html, andreport.xlsx; - concordance, discordance, and metric-distribution summaries;
- figures and provenance needed to understand the run.
Read Worked Examples for step-by-step commands and 100-sample E. coli case study for the results.