Before any lab work, there is a test that costs nothing, takes about an hour, and tells most people more than a consumer genome report will. It is a set of questions about your relatives. Medicine calls the result a pedigree. Plenty of people have drawn a family tree for a school biology class, or out of curiosity about who their ancestors were. Almost nobody has been asked for the medical version.
That is a strange gap. Sequencing gets cheaper every year and we write about it constantly. Yet the first thing a clinical geneticist does with a new patient is an interview, and it sounds nothing like a lab. Who is in your family. What were they diagnosed with, and how old were they when it happened. What did the ones who died die of, and when. The genome comes later, and it is read in the light of those answers.
What a pedigree actually is
A pedigree is a structured drawing of at least three generations: you, your siblings and children, your parents and their siblings, your grandparents. Squares for men, circles for women, a filled shape for a diagnosis, an age written beside it. Lines connect partners to children, so a condition can be followed as it moves down and across a family.
It looks too simple to be worth much. It is one of the most information-dense documents in medicine, because every filled shape is the result of a completed experiment. A relative who developed a condition did so with a real genome, in a real environment, over a real lifetime. Nothing had to be predicted. It happened, and it was counted.
Every filled shape on a pedigree is a completed experiment: a real genome, in a real environment, over a real lifetime.
One thing the drawing does not do is stay finished. A pedigree changes as more becomes known about distant relatives, and a diagnosis that surfaces in a cousin years from now can change how the whole chart reads. The version you draw this afternoon is a snapshot of what your family currently knows about itself.
Why questions can outrun sequencing
Most common conditions are polygenic. Thousands of variants each contribute a small push, the pushes interact with food, environment and time, and turning a list of variants back into a risk for one person is a hard statistical problem. The honest tools say so.
A family history skips the reconstruction. You share about half your genome with a parent or sibling and about a quarter with an aunt or a grandparent. Relatives also tend to share your ancestry, and often your kitchen, your town and your habits. When a condition keeps surfacing on one side of a family, you are looking at the combined effect of shared variants and shared life, already integrated, in the only system that counts: people.
Medicine has trusted this for decades. A first-degree relative with early colorectal cancer moves your recommended screening age forward[1]. A strong history of breast cancer changes when imaging starts, and can justify genetic testing that would otherwise never be offered[2]. None of this waits for a sequencer. The pedigree alone moves the decision.
The patterns a geneticist looks for
A pedigree is read for shape more than for any single diagnosis. How early, how often, and on which side.
Age at onset first. A grandmother with breast cancer at 78 and one with breast cancer at 41 are different findings. A common condition arriving decades ahead of schedule is the strongest single signal a family history can carry.
Then repetition on one side. Two or three relatives with related conditions in the same bloodline suggest something travelling with it. The same count scattered across both sides of the family usually means far less.
Then the recognised pairings. Some conditions cluster because one variant raises the risk of both: breast with ovarian cancer[2], colorectal with endometrial[3], melanoma with pancreatic[4]. A family showing one of each may be describing a single cause.
And the rare events. Male breast cancer. Ovarian cancer at any age. A sudden cardiac death at 44 that nobody can explain. Some diagnoses are uncommon enough that a single appearance earns attention on its own.
The shapes of inheritance
The pattern of who is affected carries information of its own. A condition that appears in every generation, in both sexes, reaching roughly half the children of anyone affected, is behaving like a dominant variant. One that skips generations and surfaces in siblings whose parents are fine is behaving like a recessive one, and it becomes more likely where the parents share ancestry.
A condition that keeps affecting men while travelling through unaffected mothers points at the X chromosome. One passed by mothers to all their children, and by fathers to none, points at the mitochondria, the one part of your genome inherited from a single parent. A geneticist can often name the likely mode of inheritance from the drawing alone, before any molecule is examined.
What the pedigree misses
Family history has real limits, and pretending otherwise would repeat the mistake the dashboards make. Small families hide patterns: a variant cannot show itself in relatives who were never born, or who died young of something else. An uneventful history in a family of three is weak evidence of anything.
Memory is worse than people expect. Diagnoses blur across a generation of retelling; ovarian cancer becomes a stomach problem, an arrhythmia becomes a heart attack. Relatives lose touch, records get sealed, adoptions go unmentioned. And many people who carry a high-risk variant have a family history that looks completely ordinary[5], because the variant travelled through carriers who stayed lucky, or arrived new with them.
Some of what is missing was never forgotten. A relative who regards a condition as shameful may simply not mention it, and a family that is not on speaking terms does not pass information across the gap. The diagnoses that go undisclosed tend to be predictable ones: who died of cancer, whose son has an intellectual disability. So the informant decides what the chart can show. The same family can produce a pedigree full of signal, or one that reads as a completely healthy family, depending on who agreed to sit down and what they were willing to say.
So the pedigree and the sequence answer different questions, and neither replaces the other. The family history says where to look and how hard. The genome, read with that context, can say whether the suspicion has a molecular explanation. The order matters: questions first, sequencing second.
Why nobody asks
If the tool is this good, why has no one ever asked you? Partly time: an average appointment cannot fit a three-generation interview, so family history collapses into a checkbox. Partly fashion: sequencing feels like the future and interviews feel like the past, so the attention follows the machines. And partly because nobody owns the task. Your doctor assumes you would mention anything important. You assume they would ask if it mattered.
The result is that the highest-yield genetic instrument in medicine sits unused in most families, while the same families are sold chip tests that read a fraction of a percent of the genome and attach no questions at all.
How to take one
Most of a pedigree can be assembled in an afternoon, and the best time is while the older generation can still be asked. For every relative across three generations, write down four things: what they were diagnosed with, at what age, what they died of, and at what age. Keep the two sides of the family separate, because a pattern confined to one line means more than the same pattern spread across both.
Add ancestry, since some variants concentrate in particular populations. Mark adoptions and estrangements, so gaps read as gaps rather than as good news. Precision beats completeness: breast cancer at 42 is worth ten entries of cancer, not sure which.
Then treat the document as alive. A cousin diagnosed this year changes your pedigree, and possibly your screening schedule, in a way your genome never will. The sequence you were born with is fixed; the evidence about what it does keeps arriving. There is even a modest tradition here: since 2004 the US Surgeon General has designated Thanksgiving, the one day the whole family is in a room, as National Family Health History Day[6]. Almost nobody knows that. The idea is still right.
Questions first
There is a reason our clinical geneticist starts every conversation with family history and only opens the genome afterwards. The pedigree sets the questions. The sequence, where it helps, refines the answers. Run in that order, a genome becomes an instrument pointed at things that plausibly matter to you. Run in reverse, it becomes a list of maybes with nobody attached.
Somewhere in your family there was already a consultation about you. It happened in a hospital corridor, or over a phone call nobody wrote down, when a relative was diagnosed and the rest of the family quietly wondered what it meant for them. A pedigree is just that conversation, finished properly. Most families have started one. Almost none have written it down.