Skip to content

Input Files

HOMERun discovers samples from FASTQ filenames. It does not use a sample sheet.

Filename pattern

Each filename must contain, in order:

  1. a two-token species name;
  2. an optional sample name;
  3. optional condition tokens;
  4. an assay token: csRNA, sRNA, totalRNA, or RNA;
  5. a lowercase replicate marker such as r1, r2, or rep1;
  6. optional lane, accession, and read metadata.

Hyphens and underscores are treated as separators. The Illumina read markers R1 and R2 are uppercase and are not interpreted as replicate markers.

Examples:

Filename Species Sample Replicate identity
homo_sapiens_K562_csRNA-r1_DB422_S1_R1_001.fastq.gz homo_sapiens K562 csRNA_r1
homo_sapiens_K562_D2_Post_csRNA_r2_R1.fastq.gz homo_sapiens K562 D2_Post_csRNA_r2
Apis_mellifera_csRNA_r1_R1.fastq.gz apis_mellifera apis_mellifera csRNA_r1

When no distinct sample token exists, HOMERun reuses the species name as the sample name.

Placement

FASTQs may start in either location:

  • directly in the project root; --stage-raw and submit_array.sh move them into the correct species directory;
  • in Species/RawData/, ready for discovery.

The --copy-src option accepts a quoted shell glob and copies matching FASTQs during preparation. Quote the glob so the controller, rather than your current shell, receives it.

submit_array.sh \
  ... \
  --copy-src '/data/run42/*.fastq.gz'

Paired total-RNA mates are matched by parsed species, sample, condition, assay, and replicate identity. Their download accessions do not need to match.

Validate names before a long run

A filename without a recognized assay or lowercase replicate marker is rejected. Run homerun --project PROJECT --stage-raw first and inspect the resulting Species/RawData/ directories.