Genotyping vs sequencing: what each finds, and what our lab uses

Genotyping is a lab method that checks your DNA at specific, known positions, picked before the test is run.^[2] The result at each position is your genotype, often written as two DNA letters, such as CC, CT or TT.^[9] There are two letters because you usually have two copies of each gene, one from each parent.^[10]

Because it looks only where it was designed to look, genotyping can’t find a variant that isn’t on its list. Here is how it compares with other ways of reading DNA, and why our tests use it.

Your DNA is a long string of four chemical letters. Labs read it in different ways, and the way they read it decides what they can find.

• Genotyping checks specific, known positions chosen in advance, and reports which version you carry at each one. It is like checking a book for a list of known typos.
• Sequencing determines the order of the letters along a stretch of DNA.^[1] Reading a whole gene this way is like proofreading every page, so it can also find a change nobody listed in advance.

Two common lab approaches check a panel of chosen positions:

Both check only the positions they are designed for. That is their strength and their limit. The fluorescent-probe method, for example, was described in 1999 as already in use for typing variants in genes that process medicines.^[3]

Sequencing determines the order of DNA’s four chemical letters.^[1] A clinical test that sequences a panel of genes reads each gene letter by letter, so it can find rare variants and ones that have never been seen before, as well as known ones.

Letter-by-letter reading can still miss bigger changes, where a whole piece of a gene is missing or repeated. Labs look for those with a separate test. In a study of 48,456 people tested for BRCA1 and BRCA2, about 1 in 10 of the harmful changes found in the higher-risk group were these large rearrangements. The authors concluded that testing for them alongside full-gene sequencing is an appropriate approach.^[4]

SNP chips are a kind of genotyping that checks a very large number of common positions at once. They are used in big research studies and some consumer DNA tests, and they work well for common variants. In 49,908 UK Biobank participants, chip results for 108,574 common variants agreed with sequencing more than 99% of the time.^[5]

Rare variants are another story. For the rarest variants, found at a frequency below 1 in 100,000 in UK Biobank, only 16% of the chip results were confirmed by sequencing. For rare harmful BRCA1 and BRCA2 variants, the chips found 34.6% of the ones that were there, and only 4.2% of the ones they reported were real. The authors concluded that SNP chips are extremely unreliable for very rare harmful variants and shouldn’t be used to guide health decisions without validation.^[5]

Our genetic tests use genotyping. They check known variants chosen in advance, and none of them reads genes in full.

Genotyping suits variants that are well known and well studied, such as many in the genes that process medicines.^[3] How pharmacogenomics works explains how those variants become a result.

It is also why we are careful about cancer. There are hundreds of BRCA variants that could increase cancer risk,^[6] and each is individually very rare.^[5] So anyone with a personal or family history of cancer should ask their clinician about clinician-ordered full-gene testing. GeneCancer checks selected variants across 108 genes, including BRCA1 and BRCA2, by genotyping. It isn’t a full-gene test: it can miss variants that full-gene testing would find, and a result with no variant found doesn’t rule out inherited risk. If it finds a variant, ask your clinician to confirm it with a clinical test.

Selected variants only, by genotyping. Not a full-gene test.

Three terms come up when experts judge a genetic test:^[7]

• Analytical validity: how well the test tells whether a particular gene or genetic change is there or not.
• Clinical validity: how well the change it looks for is related to a disease, or to the risk of one.
• Clinical utility: whether the result gives helpful information for decisions about diagnosis, treatment or managing a disease.

Every lab that tests samples for health reasons, genetic tests included, must meet federal quality standards called CLIA. CLIA standards don’t address a test’s clinical validity or clinical utility.^[7]

A lab that develops its own test, a laboratory-developed test, must establish how the test performs before it reports any results: how often it gets the right answer, how repeatable it is, how well it detects what it is meant to, and what else might interfere.^[8] Our lab page lists our laboratory’s CLIA certificate.

1. Which method does it use? Genotyping, a chip or sequencing, and for which genes.
2. Which variants does it check? For a test of selected variants, ask what is on the list, and what isn’t.
3. Does it look for missing or repeated pieces of genes? Letter-by-letter reading alone can miss them.^[4]
4. Is the lab CLIA-certified, and was this test validated there?
5. What happens if it finds a rare, serious variant? Ask whether the lab recommends confirming it with a clinical test.
6. How does it report uncertain results?
7. Who can you talk to about your result?

• Anything about positions a test didn’t check. A genotyping result speaks only for the variants on its panel.
• Whether a rare finding is real, on its own. Rare variant calls should be confirmed before anyone acts on them.^[5]
• What to do. A result is information to take to your clinician, not a decision.