A field guide to genetic testing
"Genetic testing" covers everything from a $99 spit kit to a genome read in a day for a baby in intensive care. They look at different things, miss different things and answer to different rules. Here's a map, with a tool for matching the test to the question.
By Reid Robison, MD. 2,847 words, with a map of what each test can see, how testing for trisomies and for developmental differences has changed, the race to read a newborn's genome, who checks the tests, and a test chooser. Part 2 of 2; part 1 is I finally read my own genome.
How much of the genome each test reads, not to scale at the low end. An exome is about 2 percent of the letters, but it holds most of the known disease-causing changes.
- Start with the question, not the test. "Why is my child not talking yet?" and "What's my ancestry?" need completely different tests.
- Consumer chips aren't sequencing. They read a few hundred thousand chosen spots. In one lab's review, 40 percent of the health variants people brought in from raw chip data were wrong.1
- For children with developmental differences, an exome or genome is now a first-line test. It finds a cause about a third of the time, roughly double the older approach.23
- Prenatal blood tests for trisomies are screening, not diagnosis. A positive result needs confirming, and how often it's right depends on the mother's age.4
- A critically ill newborn can now be sequenced in days, sometimes hours. In one California program, 40 percent of babies got a diagnosis and a third had their care changed.5
- Nobody checks everything. Lab regulators check that a test is run correctly. Whether a lab's own test means what it claims mostly goes unchecked; the FDA's attempt to change that was struck down in 2025.6
1One page at a time
Think of a genome as a book of about 2 million pages, and a genetic disease as a single typo somewhere in it. For most of my career, testing meant guessing which page the typo was on and paying to read that one page.
Years ago I saw a young girl for a genetics consult. She couldn't speak, she had other developmental problems, and she had already been from doctor to doctor. I ordered two single-gene tests for the two conditions I thought most likely. Many weeks and several thousand dollars later, both came back negative. We'd read two pages and found nothing, and we were no closer to an answer.
A while later I met a two-year-old with a feeding tube who couldn't hold her head up. She had a seizure in front of me. This time we sequenced every gene at once and found one wrong letter in a sodium channel gene. It explained everything. Her parents found other families with the same condition that night, and they learned about a seizure medicine she hadn't tried. Before she started it, I went back into her genome to check that she didn't carry a variant that raises the risk of a dangerous rash on that drug.
That's the shift in one comparison: from guessing pages to reading the book, then going back to it as new questions come up. But not every test reads the book, and even a whole genome misses a few kinds of typos. So here's the map.
2What each test can see
Genetic changes come in different sizes, from one wrong letter to a whole extra chromosome. Each test is built to see some sizes and is blind to others. Pick a test to see its strengths and blind spots.
What can this test see?
Six kinds of genetic change, from smallest to largest, plus repeats. Filled means the test sees it well, half means partly or only in some labs, empty means it doesn't.
Pick a test above, or .
The pattern to notice: no single test sees everything. A short-read genome comes closest, but repeat expansions like fragile X, and some rearrangements, can still slip past it. That's why a child might get a genome and a separate fragile X test, and why "my genome was normal" is never quite the same as "I don't have a genetic condition."
3Chips aren't sequencing
Most consumer DNA tests, including the big ancestry and health kits, use a genotyping chip. Instead of reading the book, a chip checks a few hundred thousand preselected spots, each one a place where people commonly differ. That's excellent for ancestry and for common traits. It's much weaker for rare disease, because most disease-causing changes are rare and so aren't on the chip.
Two consequences matter. First, a negative result on a chip means very little. When the FDA authorized the first consumer BRCA test in 2018, it checked three variants, the ones most common in people of Ashkenazi Jewish descent. There are more than 1,000 known harmful BRCA variants.7 Second, the rare variants a chip does report are often wrong. When one clinical lab retested variants that people had found in their raw chip data and run through third-party interpretation sites, 40 percent of them weren't there at all.1
So my advice on consumer kits is simple. Enjoy the ancestry. Treat any health finding as a lead, not a result, and confirm it in a clinical lab before anyone acts on it. And know that most consumer genetics companies aren't covered by HIPAA, the medical privacy law, so their privacy policy is your main protection. I covered the rest of the privacy picture in part 1.
4Children who develop differently
When a child has developmental delay, intellectual disability, autism with other features, or birth defects, families often want to know why. A genetic answer can end years of searching, point to the right treatment or specialist, predict what's ahead, and tell parents the chances for their next child.
How we look for that answer has changed three times in my career. For decades the first test was a karyotype: chromosomes stained and photographed under a microscope, which can spot an extra or missing chromosome or a large chunk out of place. Then came the chromosomal microarray, which finds missing or extra pieces far too small to see under a microscope. Now an exome or genome comes first.
How often each approach finds an answer
Share of children with neurodevelopmental conditions who get a genetic diagnosis, by first test.
Exome yield was 31 percent for children with developmental differences alone and 53 percent when they also had other medical conditions, in a pooled analysis of 30 studies.2 Karyotype figure from the 2010 consensus statement that made microarray the first test.8
In 2021 the American College of Medical Genetics recommended exome or genome sequencing as a first- or second-line test for children with congenital anomalies, developmental delay or intellectual disability.3 A genome has an added advantage: it can usually pick up the missing and extra pieces a microarray finds, so one test can do the work of two. Fragile X, the most common inherited cause of intellectual disability, is the exception. It's caused by a repeat that grows over generations, and most labs still test for it separately.
Two-thirds of children still don't get an answer on the first pass. That isn't the end. The data can be reanalyzed as new genes are discovered, often for free or a modest fee, and a year or two later it sometimes finds what it missed.
5Trisomies, then and now
Down syndrome is the clearest example of how far prenatal testing has come. It's caused by a third copy of chromosome 21, which is why it's also called trisomy 21. Tap through the history.9
A short history of finding trisomy 21
Tap a year.
Today's blood test, called cell-free DNA screening, reads tiny fragments of placental DNA in the mother's blood and counts whether there's too much chromosome 21. It catches about 99 percent of pregnancies with trisomy 21 and rarely raises a false alarm. But it's still a screening test, and that matters most when a result comes back positive.4
Same test, different odds
If cell-free DNA screening comes back positive for trisomy 21, how likely is it that the baby actually has it? It depends on how common trisomy 21 is at the mother's age. Pick an age.
Approximate published estimates; exact figures vary by lab, gestational age and the population studied.10
That's why a positive screen is followed by a diagnostic test, either chorionic villus sampling or amniocentesis, which looks at the baby's cells directly. Professional guidelines now say screening and diagnostic testing should be offered to every pregnant patient, whatever her age, and that choosing no testing at all is also a reasonable option.4
6A genome in a day
Every baby born in the US already gets a genetic screen of sorts: the heel-prick blood spot taken in the first days of life. It began in the 1960s with a test for phenylketonuria, a condition where a special diet prevents severe intellectual disability. The national recommended list now has 40 core conditions.11 Most of them are found by measuring chemicals in the blood, not by reading DNA.
The bigger change is in the intensive care unit. Some critically ill newborns have a genetic condition, and for them time is everything: the right diagnosis can mean a specific treatment, stopping a useless one, or sparing a family a long search.
- Speed. In 2022 a Stanford team went from blood sample to genetic diagnosis in about 7 hours in its fastest cases, using a new kind of sequencer and cloud computing.12
- Results. In Project Baby Bear, California funded rapid genomes for 184 critically ill infants in five hospitals. Seventy-four (40 percent) got a diagnosis, with results in a median of three days, and 58 (32 percent) had their care changed.5
- Cost. The same program estimated $2.2 to 2.9 million in savings on hospital care, against about $1.7 million spent on testing and the care it prompted.5
The next question is whether every newborn should be sequenced, sick or not. In New York, the GUARDIAN study offered genome sequencing to healthy newborns alongside standard screening, and 72 percent of families agreed. In its first 4,000 babies, the genome flagged about twice as many treatable conditions as standard screening, once a common enzyme deficiency was set aside.13 In England, the Generation Study aims to sequence 100,000 newborns for more than 200 conditions.14
I'm hopeful and careful at the same time. Finding a treatable condition before symptoms is the whole point of newborn screening. But a genome also turns up variants whose meaning is unclear, and every false alarm lands on new parents. The studies are asking the right questions, and I want their answers before this becomes routine.
7Who checks the tests
When I got the FDA's letter in 2016, the question was who gets to decide what a genetic test may tell you. Ten years later the honest answer is still: it depends on the test.
Clinical labs are certified under CLIA, a federal law overseen by Medicare's agency, and many are also accredited by the College of American Pathologists. This checks that the lab measures what it says it measures, reliably. It doesn't check whether the result means what the report says it means.
That's the FDA's question. It reviews test kits sold to labs and tests sold straight to consumers, like the 23andMe health reports. But most clinical genetic tests are lab-developed tests, designed and run in a single lab, and the FDA has long mostly left those alone. It finalized a rule in 2024 to start reviewing them. A federal court struck it down in March 2025, and the FDA formally rescinded it that September.6
HIPAA protects results held by your doctors, hospitals and the labs that bill insurance, but not most consumer genetics companies. GINA limits health insurers and employers, with gaps covered in part 1. States add their own rules, unevenly.
In practice, that leaves professional guidelines, genetic counselors and the reputation of the lab doing much of the work regulators don't. It's one reason I'd choose a lab that publishes its methods and its accuracy, and one reason a genetic counselor is worth the visit.
8Which test fits your question?
Here's how I'd match a question to a starting test. It's a starting point for a conversation with a doctor or genetic counselor, not a substitute for one.
Start with your question
Pick the one closest to yours.
Whatever the test, a few questions are worth asking before the sample goes in. What exactly will it look at, and what will it miss? What happens if it finds something nobody asked about? Who will help us understand the result? And who else might see it? If no one can answer those, it's worth finding someone who can. The National Society of Genetic Counselors keeps a directory, and many counselors now see patients by video.
I spent years trying to get genomes into people's hands, and this month I finally read my own. What I keep coming back to is that the technology is the easy part now. The hard part is matching the right test to the right question and helping people understand what came back.
Disclosures. Reid Robison co-founded Tute Genomics and was its CEO until its acquisition in 2016; he has no financial interest in any genomics company. In the past three years he has received consulting or speaker fees from Beckley Psytech, Definium Therapeutics (formerly MindMed) and Janssen (Johnson & Johnson), and research funding paid to his research sites from many sponsors, listed on the disclosures page. This post is educational and isn't medical advice. The patient stories are drawn from his practice, with details changed or left out to protect privacy.
How this post was made
I use AI tools in building this site, and I'd rather be open about how. The questions, the clinical stories and the judgment calls are mine. Claude helped research the numbers, draft the text and build the interactive pieces. I checked the figures against the sources below. Any errors are mine.
References
- Tandy-Connor S, Guiltinan J, Krempely K, et al. False-positive results released by direct-to-consumer genetic tests highlight the importance of clinical confirmation testing for appropriate patient care. Genet Med. 2018;20(12):1515-1521. doi:10.1038/gim.2018.38
- Srivastava S, Love-Nichols JA, Dies KA, et al. Meta-analysis and multidisciplinary consensus statement: exome sequencing is a first-tier clinical diagnostic test for individuals with neurodevelopmental disorders. Genet Med. 2019;21(11):2413-2421. doi:10.1038/s41436-019-0554-6
- Manickam K, McClain MR, Demmer LA, et al. Exome and genome sequencing for pediatric patients with congenital anomalies or intellectual disability: an evidence-based clinical guideline of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(11):2029-2037. doi:10.1038/s41436-021-01242-6
- American College of Obstetricians and Gynecologists. Screening for fetal chromosomal abnormalities. ACOG Practice Bulletin No. 226. Obstet Gynecol. 2020;136(4):e48-e69. doi:10.1097/AOG.0000000000004084
- Dimmock D, Caylor S, Waldman B, et al. Project Baby Bear: rapid precision care incorporating rWGS in 5 California children's hospitals demonstrates improved clinical outcomes and reduced costs of care. Am J Hum Genet. 2021;108(7):1231-1238. doi:10.1016/j.ajhg.2021.05.008
- Federal district court vacates FDA's laboratory developed tests final rule (March 31, 2025). FDA Law Blog. thefdalawblog.com; College of American Pathologists. FDA LDT rule rescinded. 2025. cap.org
- Breastcancer.org. FDA authorizes 23andMe BRCA genetic test, but test only identifies three mutations. March 2018. breastcancer.org
- Miller DT, Adam MP, Aradhya S, et al. Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies. Am J Hum Genet. 2010;86(5):749-764. doi:10.1016/j.ajhg.2010.04.006
- Global Down Syndrome Foundation. Down syndrome research and medical care timeline. globaldownsyndrome.org
- Positive predictive value of cell-free DNA screening for trisomy 21 by maternal age, as summarized in FPnotebook, Placental cell-free DNA. fpnotebook.com
- Health Resources and Services Administration. Recommended Uniform Screening Panel, as of December 2025. hrsa.gov
- Gorzynski JE, Goenka SD, Shafin K, et al. Ultrarapid nanopore genome sequencing in a critical care setting. N Engl J Med. 2022;386(7):700-702. doi:10.1056/NEJMc2112090
- Ziegler A, Koval-Burt C, Kay DM, et al. Expanded newborn screening using genome sequencing for early actionable conditions. JAMA. 2025;333(3):232-240. Published online October 24, 2024. doi:10.1001/jama.2024.19662
- Isle of Wight NHS Trust. Research study testing newborn babies for over 200 rare genetic conditions launched. January 21, 2026. iow.nhs.uk