Project layout¶

Omnomnomics keeps inputs, intermediate files, results, and provenance in one experiment directory.

EXPERIMENT_DIR/
├── FASTQ/
├── trimmed_FASTQ/
├── fastqc_reports/
├── BAM/
├── filtered_BAM/
├── BigWigs/
├── merged_hubs/
├── peak_calling/
├── DE_calling/
├── MultiQC/
├── run_configs/
├── run_logs/
└── slurm_logs/

Not every assay or selected stage range creates every directory.

Directory	Purpose
FASTQ	original compressed reads; retained by all retention policies
trimmed_FASTQ	adapter-trimmed reads and trim metrics
BAM	aligner output, lane-merged BAMs, and mapper statistics
filtered_BAM	sorted, filtered BAMs, indexes, and BAM QC
BigWigs	per-sample browser signal tracks
merged_hubs	grouped UCSC track hubs
peak_calling	ATAC/ChIP feature definitions and pre-DE reports
DE_calling	count matrices and differential-analysis results
run_configs	fully resolved per-run configuration
run_logs	run provenance, stage summaries, and cached lightweight metrics
slurm_logs	controller and worker standard output grouped by rule

The reference root is separate from experiment data. A normalized assembly has fasta/genome.fa, annotation/genes.gtf, aligner index directories, and aux/.