Precision medicine can feel like a black box: you give a sample, and weeks later a report names a gene and a verdict. This explainer walks the workflow end to end — from the clinic conversation, through the laboratory steps, to the five-tier report a clinician reads with you. Knowing each stage helps you ask better questions and understand why a result can be a clear answer, a maybe, or nothing at all. It draws on the same registry tradition behind PROMPT, the hereditary-cancer registry whose patient-reported variant data has appeared in ASCO/JCO literature.
Step 1: The clinic conversation and consent
Testing should begin before any swab. A clinician or genetic counsellor reviews your personal and family history against published criteria — for hereditary breast and ovarian cancer, for example, NCCN testing criteria and the ASCO germline-panel-selection guidance shape who is offered which test. This stage sets expectations: what the test can and cannot find, the chance of an uncertain result, and what a positive finding might mean for relatives.
Step 2: Sampling and the kit
Most germline (inherited) testing uses blood or a saliva collection kit such as an OG-500 tube; tumour (somatic) testing uses a tissue biopsy. The distinction matters: a germline test reads the DNA you were born with and inherited, while a somatic test reads mutations that arose in a tumour. A pharmacogenomic finding — say a CYP2C19 loss-of-function allele that affects how you activate clopidogrel — is germline, even though it can sometimes be salvaged from tumour sequencing data.
Step 3: Sequencing and coverage depth
In the lab, DNA is fragmented, enriched for the regions of interest, and read on a next-generation sequencer. “Depth” — how many times each base is read — is a core quality metric. For germline variant calling, laboratories typically target at least 50x and prefer around 100x; clinical whole-exome sequencing commonly runs at roughly 100–200x average depth, with the great majority of bases covered well above 20x. Targeted panels concentrate reads on a short gene list, so they often reach higher, more uniform depth than an exome at the same sequencing effort — useful when sensitivity for a known set of genes matters most.
- Targeted panel — a curated gene list (e.g.
BRCA1,BRCA2,PALB2,ATM,CHEK2,TP53); deep, uniform coverage. - Whole-exome — all protein-coding regions; broad but shallower per base.
- Whole-genome — coding and non-coding DNA; broadest, most data to interpret.
Step 4: Variant calling and ACMG classification
Software compares your reads to a reference genome and lists differences — variants such as rs80357906. Each clinically relevant variant is then classified using the 2015 ACMG/AMP framework, which weighs evidence (population frequency, computational predictions, functional studies, segregation) into five tiers. “Likely pathogenic” and “likely benign” denote greater than ~90% certainty in either direction.
| ACMG tier | Plain meaning |
|---|---|
| Pathogenic | Disease-causing |
| Likely pathogenic | Probably disease-causing (>90% certainty) |
| Variant of uncertain significance (VUS) | Not enough evidence to call either way |
| Likely benign | Probably harmless (>90% certainty) |
| Benign | Harmless |
Step 5: The report and what “yield” means
The signed report reaches your clinician. “Diagnostic yield” is the share of tested people who get an actionable positive. In hereditary breast cancer cohorts meeting NCCN criteria, multigene panels have detected germline pathogenic variants in roughly a quarter of patients, with a meaningful fraction lying outside BRCA1/BRCA2 — one reason panels, not single-gene tests, are now common. A VUS is not a positive result; it may be reclassified years later as evidence accumulates, which is why staying in touch with your clinic matters.
The bottom line
From the consent conversation to the final classification, each step shapes what your report can say. Depth determines confidence in the raw calls; ACMG tiers translate those calls into clinical language; and yield reminds us that “no pathogenic variant found” is a common, valid outcome — not a failure. Understanding the pipeline lets you read a report as the structured, probabilistic document it is.
Educational explainer, not medical advice. For your own testing or results, speak with a clinician or genetic counsellor.