Run outputs

Outputs and reports

Review the files produced for preprocessing, quantification, analysis status, and reporting.

Understand the methodsBack to overview

Preprocessing and quantification

fastp/

Processed reads plus per-sample JSON and HTML quality-control records. These files document preprocessing and do not contain differential expression results.

salmon/<sample>/quant.sf

Salmon transcript-level abundance, effective length, and estimated-count output for each sample. tximport consumes these files with the transcript-to-gene mapping.

Sample metadata

The normalized sample table links sample identifiers, conditions, and input files. For downloaded data, acquisition records store the source SRA/ENA accession.

Gene-level matrices

tximport/txi.tsv

The gene-level count matrix used as input to DESeq2. TPM is not used as DESeq2 input.

tximport/gene_tpm.tsv

Gene-level TPM when available. This is a gene-level abundance measure and is never passed to DESeq2.

Mapping QC

tximport records mapping and identifier-normalization checks used to detect transcript-to-gene compatibility problems.

DESeq2 status and results

deseq2/status.json records the selected analysis mode, sample counts per condition, and which output files were generated. Use this file instead of inferring analysis status from file presence.

  • deseq2/results.tsv contains standard differential expression columns in differential mode. In QC-only mode it has a header and zero data rows.
  • deseq2/normalized_counts.tsv contains DESeq2-normalized counts when the count matrix permits normalization.
  • PCA and sample-distance plots are produced when applicable. Inferential MA, volcano, adjusted-p-value, and fold-change plots are produced only in differential mode.
  • deseq2/qc_summary.tsv and deseq2/qc_summary.json summarize the mode and available output files.
QC-only mode does not calculate or report p-values or adjusted p-values. Inferential contrasts are inactive. The results table has no data rows, differential expression plots are marked not applicable, and enrichment is not run.

Saved run record

The run/ directory stores immutable inputs used to resume and review a run. Key records include run/config_resolved.yaml, normalized sample metadata, the analysis plan, commands and logs, reference provenance, the Git revision, and run/versions.tsv. Later GUI edits do not rewrite these run inputs.

Because the saved run inputs do not change, Snakemake can resume only the required workflow steps.

Self-contained HTML report

The final report is always located at report/report.html relative to the run directory. It summarizes sample metadata, references, fastp, Salmon, gene-level counts and TPM availability, analysis mode, QC, and applicable differential results.

The report is static and self-contained: it does not require an internet connection, external JavaScript, fonts, analytics, or other web assets to render. Keep the surrounding run directory when you also need the result files, logs, and reference provenance.

See the full output reference and GitHub repository for current file contracts.