Creating Samples

The sample is the physical specimen taken from a subject. A sample owns the sequencing files (VCF, BAM, FASTQ) and the metadata describing how the specimen was collected and processed.

Creating a sample

1. Open Samples from the left nav and click New Sample.
2. Choose the Subject this sample was taken from.
3. Enter a Sample ID — a human-readable identifier, unique within the project (e.g. NA12878-DX-01).
4. Optionally add specimen metadata: specimen type, collection date, sequencing type, and any other identifiers your workflow needs.
5. Click Create Sample.

Creating a sample requires a project Contributor or Admin role.

Files

Sample files live under samples/{slug}/uploads/, where {slug} is derived from the Sample ID — see File Paths:

File	Purpose
*.vcf.gz	Variant calls — the input for tertiary reporters and the variant browser.
*.bam (+ .bai)	Aligned reads, useful for IGV inspection.
*.fastq.gz	Raw reads, when the pipeline does alignment.

Tabix index (.tbi) is only needed for order VCFs

A VCF that goes into an order (for example, after a completed analysis) must be bgzipped and tabix-indexed (.vcf.gz + .tbi) so the variant browser can build its cache. A VCF you upload as the starting input to a workflow does not need a .tbi index.

Where things live in the UI

• Samples list — /samples. Browse and search the project's samples.
• Sample detail — /samples/{sample_id}. View the sample's metadata, its subject, the orders it participates in, and its uploaded files.
• New sample — /samples/new. Create a sample by choosing a subject, entering a Sample ID, and adding optional specimen metadata.

Related reading

• Subjects — where samples come from.
• Orders — how samples participate in a case.
• For scripted creation, manifest upload, and samplesheet generation, see CLI → Managing Samples.
• Analyses — how a sample's files feed a pipeline.