Daniel sent us this one — he's asking, in simple terms, what DNA actually is. Then he's got this layered follow-up: does our DNA change over our lifetime, and when a couple conceives, is the child getting a snapshot of the father's DNA at that exact moment, or are these fixed expressions passed down? And then he tacks on a historical curveball — what are the most conspicuous milestone years in DNA research, and when was it first discovered? It's basically four questions stacked in a trench coat, pretending to be one prompt.
They're good questions. The third one especially — the "snapshot or fixed expression" framing — that's the kind of thing people assume they know the answer to, and then you actually think about it for thirty seconds and realize there's real depth there.
Where do we start? What is DNA, simple terms, no textbook preamble.
DNA is a string of four chemical letters — A, T, C, and G — that stores instructions for building and running a living organism. About three billion of these letters in the human genome, arranged in a specific order across twenty-three pairs of chromosomes. If you stretched the DNA in a single human cell end to end, it would be about two meters long. All coiled into a nucleus about six micrometers across.
Two meters of string stuffed into something smaller than a dust mite. That's already absurd.
The packing mechanism is genuinely one of the most underappreciated engineering feats in biology. Histone proteins act as spools, DNA wraps around them like thread, and then that whole structure gets folded and refolded at multiple scales. It's like taking a rope the length of a car and cramming it into a poppy seed without tangling.
Somehow the cell still manages to find and read specific sections on demand.
That's the transcription machinery — proteins that unspool just the right segment, copy it into RNA, and then let it spool back up. The access problem is harder than finding a specific sentence in a library where every book is simultaneously being used as furniture.
That's the "what." Now the "does it change" question. I've heard both answers — yes and no — depending on who's talking and what axe they're grinding.
The short answer is yes, DNA changes over a lifetime, in multiple distinct ways. And this is where most popular science coverage collapses a bunch of different phenomena into one blur. Let me separate them. First, somatic mutations — these are changes to the DNA sequence in individual cells, acquired as you age. Every time a cell divides, it has to copy its entire genome, and the copying machinery makes errors. The error rate is extraordinarily low — about one mistake per hundred million base pairs per division — but you have trillions of cells and some of them divide constantly.
It's a numbers game. Low error rate times astronomical number of divisions equals a meaningful accumulation.
Skin cells, intestinal lining cells, blood stem cells — they're all dividing throughout your life, and each division is a chance for a typo. By the time you're sixty, a typical epithelial cell might carry a few thousand mutations that weren't there at birth. Most are harmless — they land in non-coding regions or don't change protein function. But some can be drivers for cancer, which is why cancer incidence rises with age. It's not that aging causes cancer directly — it's that you've accumulated more rolls of the dice.
That's somatic cells. What about the ones that matter for the inheritance question — the germline?
That's the second layer, and it's the one directly relevant to the "snapshot" question. Sperm and egg cells are made through a process called meiosis, and the DNA in those cells does not remain frozen from birth. In men, sperm-producing stem cells divide continuously throughout life — about twenty-three divisions per year. Each division introduces the same tiny risk of a copying error. So a man's sperm at age forty carries, on average, about twenty-five to thirty more new mutations than his sperm at age twenty-five.
That's a real number. Twenty-five to thirty new mutations that weren't there when he was younger.
The research backs this up. A landmark study in Nature in twenty twelve, led by Kári Stefánsson at deCODE Genetics in Iceland, sequenced the genomes of seventy-eight parent-child trios and found that a father passes on roughly two new mutations for every year of his age at conception. A twenty-year-old father passes about twenty-five new mutations to his child. A forty-year-old father passes about sixty-five. The mother's age has almost no effect on the mutation count — her eggs were formed before she was born and sit in a kind of suspended animation.
The child gets the father's DNA as it existed at the moment of conception — typos and all — not some pristine original copy from his birth.
It IS a snapshot — a high-fidelity snapshot of the father's germline DNA at that moment in time, complete with the accumulated scribbles of a lifetime of cell divisions. And that's just the sequence changes. There's a third layer that's even more dynamic: epigenetics.
Epigenetics — the dimmer switch, not the wiring.
I like that. Chemical tags — primarily methyl groups — that attach to DNA and change how genes are expressed without changing the underlying sequence. And these tags change constantly throughout life. Diet, stress, exercise, smoking, toxin exposure — all of these can add or remove epigenetic marks. Some are stable for years. Some cycle daily with circadian rhythms.
Can those be passed to children too?
Mostly no — there's a massive epigenetic reprogramming event that happens shortly after fertilization. Most of the parental methylation marks get wiped and reset. But not all. There's a phenomenon called genomic imprinting where certain genes escape the wipe and retain parental epigenetic marks. And there's growing evidence from animal studies that some environmentally induced epigenetic changes can persist across generations, though the mechanism is debated and the effect size in humans is unclear. I'm not sure about this part, but the best evidence suggests it's real but subtler than the headlines claim.
"Epigenetic inheritance" sounds like the kind of phrase that launches a thousand wellness podcasts and approximately zero controlled trials.
That's unfair but not entirely wrong. The field has a reproducibility problem. But the core phenomenon — that DNA sequence accumulates mutations and epigenetic marks shift — is rock solid. So to answer the prompt directly: yes, DNA changes. In somatic cells, it mutates and accumulates epigenetic drift. In the germline, it mutates at a predictable rate, and those mutations are faithfully transmitted. The child gets the father's DNA as it stood at the moment of conception, not some Platonic ideal of his genome.
Which means, in a very literal sense, that older fathers are passing down slightly different genetic material than they would have if they'd had the same child ten years earlier.
Those mutations aren't just neutral. Paternal age is correlated with increased risk of certain conditions in offspring — autism spectrum disorder and schizophrenia are the two most studied. A twenty sixteen study in JAMA Psychiatry found that children born to fathers over forty-five had about a three-and-a-half times higher risk of autism compared to fathers in their early twenties. The absolute risk is still low — we're talking about going from roughly one percent to three or four percent — but the relative increase is substantial and it's been replicated across multiple large cohorts.
That's a number that should probably be more widely known. So we've got somatic mutations, germline mutations that accumulate with paternal age, and epigenetic shifts, most of which get wiped but some don't. DNA is not a fixed blueprint. It's more like a document that's being lightly edited every single day, and then the edited version gets photocopied and handed to the next generation.
With the father's age determining how many red pen marks are on the photocopy. The mother's contribution is different — her eggs are old in a literal chronological sense, formed during fetal development, and they accumulate damage to the cellular machinery rather than new sequence mutations. That's why chromosomal abnormalities like trisomy twenty-one increase with maternal age. Different mechanism, same directional effect: older parents, more genetic risk.
Two routes to the same destination. Biology loves redundancy in all the wrong ways.
Now the fourth part of the prompt — the history. When was DNA first discovered, and what are the conspicuous milestone years?
And I'm guessing the answer to "when was it first discovered" depends on what you count as "discovered.
The substance itself — DNA as a physical molecule — was first isolated in eighteen sixty-nine by a Swiss physician named Friedrich Miescher. He was working in a lab in Tübingen, Germany, studying the chemical composition of white blood cells. He'd collect pus from used surgical bandages — I'm not making this up — and from that he extracted a phosphorus-rich substance from the cell nuclei that he called "nuclein.
Pus from bandages. The glamour of foundational science.
Miescher had no idea what it did. He speculated it might be a storehouse for phosphorus, or maybe something structural. The idea that it carried hereditary information was decades away. For the next seventy-five years, most biologists thought proteins were the genetic material — proteins are more complex, twenty amino acids instead of four bases, so they seemed like a better candidate for encoding the complexity of life.
The four-letter alphabet seemed too simple.
And that assumption held until nineteen forty-four, which is the first truly conspicuous milestone year. Oswald Avery, along with Colin MacLeod and Maclyn McCarty, published a paper showing that DNA — not protein — was the "transforming principle" that carried genetic information in bacteria. They took a pathogenic strain of pneumococcus, extracted its DNA, and showed that the purified DNA alone could transform a harmless strain into a pathogenic one.
The world yawned.
The Avery experiment is now considered one of the most important papers in biology, but it didn't win a Nobel — Avery died in nineteen fifty-five, and even though the Nobel committee later admitted they'd made a mistake not awarding him, the rules don't allow posthumous prizes. The recognition went instead to the next milestone.
Nineteen fifty-three. Watson and Crick.
Watson and Crick, with Rosalind Franklin's X-ray crystallography data — specifically Photo fifty-one — showing the double helix structure. That's the moment DNA entered the public imagination, and it's the one everyone knows. But the structure alone didn't explain how it worked. What made the paper revolutionary was the throwaway line at the end: "It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material.
One of the great understatements in scientific literature. "It has not escaped our notice.
Francis Crick later admitted they put that line in as a way of staking a claim on the functional implications without having to prove them yet. It was a scientific land grab disguised as modesty.
Died of ovarian cancer in nineteen fifty-eight, at thirty-seven, before the Nobel was awarded in nineteen sixty-two. Watson, Crick, and Maurice Wilkins shared it. Franklin's contribution was essential — her X-ray crystallography provided the physical evidence for the helical structure — but the credit was distributed unevenly, and the way Watson portrayed her in his memoir "The Double Helix" was, let's say, ungenerous. The history has been substantially revised in her favor since then.
That's the foundation. What comes next in the milestone timeline?
Nineteen sixty-one to sixty-six — cracking the genetic code. Multiple labs, but Marshall Nirenberg gets the headline for the first breakthrough: he showed that the sequence UUU codes for the amino acid phenylalanine. Over the next five years, the entire code was deciphered — which three-letter combinations correspond to which amino acids, and where the stop signals are. It's the Rosetta Stone of molecular biology. By nineteen sixty-six, you could take a DNA sequence and predict the protein it would produce.
Then a long gap before the next conspicuous year?
Nineteen seventy-seven — Fred Sanger develops chain-termination sequencing, which becomes the dominant method for reading DNA for the next thirty years. Sanger sequencing is what enabled the Human Genome Project. Sanger himself won two Nobel Prizes in chemistry — one for protein sequencing in nineteen fifty-eight, one for DNA sequencing in nineteen eighty. He's one of only four people to win two Nobels in the same category.
The Human Genome Project itself?
Launched in nineteen ninety, completed in two thousand three — ahead of schedule and under budget, which is almost unheard of for a massive government science project. It cost about two point seven billion dollars and involved thousands of scientists across twenty institutions in six countries. The announcement of the "first draft" in June two thousand was a joint press conference with Bill Clinton, Tony Blair, and the project leaders. That was a conspicuous moment — heads of state treating a biology paper like a moon landing.
Now we have people spitting into tubes and getting their genome back for a couple hundred dollars.
That's the post-genome revolution. Two thousand five — the launch of next-generation sequencing, or NGS, which parallelized the process and dropped the cost by orders of magnitude. The thousand-dollar genome became a reality around twenty fifteen. The hundred-dollar genome is here now. We're heading toward the ten-dollar genome.
The conspicuous years, to summarize: eighteen sixty-nine for the isolation of DNA, nineteen forty-four for proving it's the genetic material, nineteen fifty-three for the double helix, nineteen sixty-one-to-sixty-six for cracking the code, nineteen seventy-seven for Sanger sequencing, two thousand three for the Human Genome Project completion, and then the NGS revolution starting in two thousand five.
I'd add two more. Two thousand twelve — that's when CRISPR-Cas9 was first demonstrated as a programmable gene-editing tool, in a paper by Jennifer Doudna and Emmanuelle Charpentier. That changed everything. Before CRISPR, gene editing was possible but slow, expensive, and imprecise. After CRISPR, you could target a specific gene, cut it, and modify it with a precision that was previously science fiction. They won the Nobel in Chemistry in twenty twenty.
Twenty twenty-two — the first complete, gap-free human genome sequence was published. The original Human Genome Project had left about eight percent of the genome unsequenced — mostly repetitive regions around the centromeres and telomeres that were technically impossible to assemble with the methods of the time. A consortium called the Telomere-to-Telomere Consortium, or T two T, finally closed those gaps using new long-read sequencing technologies. The final eight percent turned out to contain about two thousand genes and a lot of structural elements we're still trying to understand.
For twenty years, we were walking around with what we thought was the "complete" human genome, and it was missing eight percent.
Biology is humbling like that. Every time we think we've got the book read, we realize we've been skipping entire chapters.
Let me pull us back to something you said earlier about the paternal age effect — twenty-five to thirty extra mutations at forty versus twenty-five. That's a concrete number, but it's small in the context of three billion base pairs. Why does it matter?
Because the mutations aren't randomly distributed in terms of their effects. Most of the genome is non-coding — it doesn't directly specify proteins — so a mutation there might do nothing. But when a mutation lands in a coding region, or in a regulatory element that controls when and where a gene is expressed, it can have outsized effects. And some genes are particularly sensitive. There's a set of genes associated with brain development that are unusually large and complex, which makes them statistically more likely to be hit by random mutations. That's one hypothesis for why paternal age correlates with neurodevelopmental conditions — the genes involved are simply bigger targets.
Bigger targets for random typos. That's elegantly simple.
There's another angle. The mutations we're talking about in sperm are de novo mutations — they appear for the first time in that sperm cell and didn't exist in the father's own body. But they exist in every cell of the child that results, because that sperm's genome becomes the template for the entire organism. So a single mutation in a single sperm becomes a whole-body mutation in the offspring.
Which loops back to the "snapshot" question. The snapshot captures not just the inherited genome, but the fresh typos introduced during sperm production.
Those typos then become part of the child's permanent record. If that child is male, he might pass them on to his own children, plus his own new set of age-accumulated mutations. The mutation rate isn't just a curiosity — it's the raw material for evolution. Every new mutation in the germline is a potential evolutionary experiment, most of them neutral or slightly harmful, a tiny fraction potentially beneficial.
When we talk about DNA being "fixed," that's wrong on both timescales. Over a lifetime, it drifts. Over generations, it accumulates changes that become permanent features of lineages.
Even calling it "fixed" at the species level is wrong when you consider horizontal gene transfer in bacteria, or the fact that about eight percent of the human genome is composed of endogenous retroviruses — ancient viral DNA that integrated into our ancestors' genomes and got passed down. We're walking around with fossilized viruses in our genetic code. Some of those viral sequences have been repurposed — there's one called syncytin that's essential for placental development. Without an ancient viral infection, mammals might not have placentas.
We are, in part, domesticated viruses.
In a very literal sense. The boundary between "our" DNA and "foreign" DNA is blurrier than we like to think.
Let me ask you something about the historical timeline. You mentioned Miescher in eighteen sixty-nine. What was the gap between his discovery and anyone even suspecting DNA had something to do with heredity?
About seventy-five years. Mendel's work on pea plants was published in eighteen sixty-six, just three years before Miescher's discovery, but it was largely ignored until it was rediscovered around nineteen hundred. Even once Mendelism was accepted, the physical nature of the "gene" was a complete mystery. The chromosome theory of inheritance — the idea that genes are located on chromosomes — was developed by Thomas Hunt Morgan and his students at Columbia in the nineteen tens, using fruit flies. They won the Nobel in nineteen thirty-three. But they still didn't know whether the genetic material was the DNA or the proteins in the chromosomes.
For decades, the smartest biologists in the world were staring at the molecule of heredity and didn't recognize it.
The reason is instructive. They were looking for complexity. Proteins are built from twenty different amino acids, which can be combined in essentially infinite sequences. DNA has only four bases. It seemed too simple to encode the complexity of a human being. It's like assuming a message must be written in a rich, expressive language, and overlooking the possibility that it's written in binary.
Four letters, but the words can be any length. The combinatorial explosion does the work.
Four to the power of N, where N is the length of the sequence, grows faster than any practical mind can grasp. A sequence of just ten bases has four to the ten — over a million — possible combinations. The human genome is three billion bases long. The number of possible sequences is larger than the number of atoms in the observable universe by an incomprehensible margin.
The simplicity was the point. A four-letter alphabet is easier to copy accurately than a twenty-letter one.
That's exactly the insight Watson and Crick had. The double helix structure immediately suggests a copying mechanism because each strand serves as a template for the other. A pairs with T, C pairs with G. Split the helix, and each half contains all the information needed to reconstruct the full molecule. It's the most elegant data storage system ever invented, and it evolved by blind trial and error over billions of years.
You mentioned CRISPR earlier — twenty twelve, Doudna and Charpentier. What exactly did they demonstrate that was new? Gene editing existed before, didn't it?
The previous generation of tools — zinc finger nucleases and TALENs — worked, but they required designing and building a custom protein for each target sequence. That took months and cost tens of thousands of dollars per experiment. CRISPR uses a guide RNA molecule to find the target sequence, and RNA is cheap and easy to synthesize. You can design a new guide RNA in an afternoon and have it synthesized for under a hundred dollars. The Cas9 protein — the scissors — is the same every time. You just swap the guide.
It turned gene editing from a bespoke tailoring operation into a print-on-demand service.
That's exactly the right metaphor. And the democratization has been staggering. High school students are doing CRISPR experiments in biology class now. The barrier to entry collapsed.
Which raises the ethical questions that everyone's been arguing about for the last decade. Germline editing, designer babies, all of that.
Those arguments got real in twenty eighteen when a Chinese researcher named He Jiankui announced he'd used CRISPR to edit the CCR5 gene in human embryos, resulting in the birth of twin girls. He claimed he was trying to make them resistant to HIV. The scientific community almost unanimously condemned him — not because the goal was bad, but because the safety data wasn't there, the informed consent process was questionable, and he'd jumped past every ethical safeguard the field had been building. He went to prison for three years.
The first CRISPR babies, and it was a rogue operation. That's not the milestone anyone wanted.
It set the field back in some ways, because it made legitimate germline research harder to fund and harder to get approved. But it also forced a global conversation that had been theoretical to become very practical. We're still having that conversation. The WHO issued a governance framework in twenty twenty-one. Multiple countries have passed laws. But the tools keep improving, and the pressure to use them — for disease prevention, if not enhancement — is going to keep building.
We've covered what DNA is, how it changes, what gets passed on, and the milestone years. Let me ask one more thing about the "changes over a lifetime" piece. Is there a practical way to measure someone's DNA changes? Can you take a snapshot at age thirty and another at age fifty and compare?
You can, and people have. There are studies that do longitudinal whole-genome sequencing on the same individuals. But it's not something you'd do clinically — it's still expensive and the interpretation is complex. What IS becoming clinically relevant is something related but different: measuring the epigenetic clock.
The epigenetic clock.
Also called the Horvath clock, after Steve Horvath, who published the key paper in twenty thirteen. Certain positions in the genome gain or lose methylation at predictable rates as people age. By measuring the methylation status at a few hundred of these positions, you can estimate someone's biological age with remarkable accuracy — within about three years of their chronological age. And when the epigenetic age is higher than the chronological age, that's associated with increased risk of age-related diseases and all-cause mortality.
You can be forty chronologically and fifty epigenetically.
And that's a bad sign. It's been linked to stress, smoking, poor diet, and a bunch of other factors. There are companies now selling epigenetic age tests direct to consumers. The science is real, but the actionable information is still limited — knowing your epigenetic age is older doesn't necessarily tell you what to do about it, beyond the standard health advice you already know.
"Your cells are aging faster than your calendar. Also, eat more vegetables.
But the research is moving fast, and epigenetic clocks are increasingly used as endpoints in anti-aging clinical trials. If you want to test whether a drug slows aging, you need a biomarker that changes faster than mortality, and the epigenetic clock is currently the best candidate we have.
The field has gone from "what is this gooey substance in pus" to "we can measure how fast you're aging by reading chemical tags on your DNA" in about a hundred and fifty years.
The pace is accelerating. The first nuclear clock was just built by physicists — that was in the news this week. The precision of biological measurement is following the same trajectory as physical measurement. We're entering an era where we can read the genome, edit it, and monitor its changes in real time. The philosophical implications are as big as the medical ones.
Which brings me back to the inheritance question. If a father's DNA accumulates mutations with age, and we can now sequence genomes cheaply, is anyone studying whether the children of older fathers have measurably different genomes than siblings born when the father was younger?
Yes, and the answer is they do. There was a twenty seventeen study that sequenced the genomes of over fifteen hundred Icelandic families — three generations — and directly measured the number of new mutations in children as a function of paternal age. The correlation was almost perfectly linear. Every additional year of paternal age adds about one and a half new mutations to the child's genome. So two siblings born ten years apart to the same father will differ by about fifteen mutations that came exclusively from the father's aging germline.
That's a small number, but they're not the same fifteen. The younger sibling might get a mutation in a brain-development gene, the older sibling might get one in a non-coding region. It's a lottery.
A lottery with a slowly increasing number of tickets. Most of the tickets are blank. A few have consequences. And the ticket count goes up every year.
That's a sobering way to think about delayed parenthood. Not judgment, just biology.
It's worth noting that the absolute risks remain low for most conditions. The paternal age effect is real and measurable, but for any individual couple, the odds of a healthy child are still overwhelmingly high. It's a population-level effect that matters for public health more than for individual decision-making.
We should probably wrap this. The prompt asked for simple terms, conspicuous years, and the snapshot question. I think we've delivered.
DNA is a four-letter chemical instruction manual that changes throughout life through mutations and epigenetic shifts. Fathers pass on a snapshot of their germline DNA at the moment of conception, complete with age-accumulated typos. The discovery timeline runs from Miescher's pus-derived nuclein in eighteen sixty-nine to the gap-free genome of twenty twenty-two, with about a dozen conspicuous milestones in between.
If you're keeping score at home, you're walking around with fossilized viruses in your genome, your biological age might not match your calendar age, and every year you age, your future children's DNA gets slightly more edited. Biology: consistently unsettling.
Now: Hilbert's daily fun fact.
Hilbert: The Ethiopian game of genna, a form of field hockey played at Christmas, takes its name from the Amharic word for "Christmas" itself — Gena, also spelled Genna or Ledet — derived from the Greek "genesis," meaning birth. The name traveled across linguistic boundaries for centuries before landing in the highlands, where it now describes both a religious holiday and a sport played with curved wooden sticks and a wooden ball.
Christmas hockey in Ethiopia is etymologically Greek.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. You can find every episode at myweirdprompts dot com. If you got something out of this one, leave us a review wherever you listen — it helps.
Until next time.
