How Ancient DNA Works

When an organism dies, its DNA immediately begins breaking down into smaller and smaller pieces through chemical processes like hydrolysis and oxidation. In ancient specimens—whether a 5,000-year-old mummy or a 40,000-year-old mammoth bone—the DNA that remains exists as tiny fragments, often just 30 to 100 genetic letters long, compared to complete chromosomes containing millions of letters in living cells. These fragments get chemically bound to minerals in bones, teeth, and sediments, which paradoxically helps preserve them even as they degrade.

To extract this degraded DNA, scientists must physically pulverize the specimen into a fine powder, typically drilling into dense bones or grinding teeth, which releases the DNA fragments from their mineral matrix. The powder is then treated with chemical solutions that break down proteins and other cellular debris while leaving the DNA fragments intact. Because ancient DNA is so scarce and fragile, researchers work in ultra-clean laboratory facilities, wearing full protective gear to prevent contamination from their own modern DNA, which would overwhelm the ancient signal.

The extraction process must also contend with environmental DNA from bacteria, fungi, and soil microbes that colonized the specimen after death. In a single ancient bone sample, more than 99% of the DNA recovered often comes from these microbial invaders rather than the original organism. Special chemical treatments and purification steps help separate and concentrate the authentic ancient DNA from this overwhelming background noise.

Ancient DNA fragments carry distinctive molecular scars from their time in the ground—cytosine bases convert to uracil, creating false "mutations" when sequenced, and the fragments' ends become ragged and chemically modified. Modern sequencing machines can't directly read these damaged molecules, so scientists first attach special molecular tags called adapters to both ends of each DNA fragment. These adapters serve as handles that the sequencing technology can grip, allowing even heavily damaged fragments to be processed.

The sequencing process generates millions of short reads simultaneously using high-throughput technologies that detect fluorescent signals as DNA bases are identified one by one. Because ancient DNA is so fragmented and sparse, researchers must sequence each sample to extreme depth—reading the same genomic regions hundreds or thousands of times—to gather enough overlapping fragments to reconstruct the original genetic sequence. A single ancient human genome might require extracting and sequencing DNA from 100 grams of bone powder, generating terabytes of raw data.

Crucially, the sequencing data reveals the characteristic damage patterns that authenticate ancient DNA and distinguish it from modern contamination. The telltale excess of C-to-T changes at fragment ends acts as a molecular signature of authenticity. Researchers actually use these damage patterns as proof that their sequences genuinely come from ancient organisms rather than from modern contamination or laboratory errors.

The millions of short DNA sequences generated from an ancient specimen arrive as a chaotic jumble, like dumping a thousand shredded books into a pile and trying to reconstruct the original texts. To solve this puzzle, scientists use powerful algorithms that compare each ancient fragment to a reference genome—a complete genetic blueprint from a closely related modern species. If studying a Neanderthal, for instance, researchers align the ancient fragments to the modern human genome, positioning each fragment where it matches best.

This alignment process must account for genuine evolutionary differences between the ancient organism and the reference genome, while also recognizing and filtering out sequences from contaminating bacteria or unrelated species. The computer algorithms calculate probability scores for each fragment's placement, keeping only those that meet strict quality thresholds. Fragments that align to multiple genome locations, or that show too many mismatches, are typically discarded as ambiguous.

The assembly becomes particularly challenging when studying species with no close living relatives, like woolly mammoths or ancient cave bears. In these cases, researchers may use more distant relatives as scaffolds (Asian elephants for mammoths, for example) or attempt de novo assembly, where overlapping fragments are stitched together without a reference—a computationally intensive process that requires exceptional DNA preservation and coverage. Even with a reference genome, gaps and ambiguities remain in the final assembly, representing regions where preservation was too poor or contamination too high.

Once an ancient genome has been assembled, scientists systematically identify positions where the ancient DNA differs from reference genomes and from modern populations. These variants—single-letter changes, insertions, deletions—form a genetic fingerprint that reveals evolutionary relationships. By comparing thousands or millions of these variant positions, researchers can determine whether an ancient specimen is more closely related to one modern population versus another, or whether it represents an extinct lineage.

The comparison process uses statistical methods that account for the fragmentary nature of ancient data and the possibility of sequencing errors. Researchers focus particularly on variable positions that differ among modern human populations or species, since these informative sites carry the strongest ancestry signals. For example, comparing a 10,000-year-old European hunter-gatherer genome to modern Europeans, Africans, and Asians reveals which present-day populations carry genetic legacy from this ancient individual.

These comparisons have revealed stunning discoveries, such as identifying that Neanderthals interbred with modern humans, contributing 1-2% of the genome of present-day non-Africans. The genetic variants shared between ancient Neanderthals and modern humans appear in specific genomic regions, proving the ancestry connection and even revealing approximately when the interbreeding occurred. Similarly, comparing ancient and modern dog genomes has traced domestication events, migration routes, and breed origins across millennia.

Ancient DNA transforms evolutionary biology from a discipline of inference into one of direct observation, allowing scientists to sample organisms from specific moments in the past rather than merely extrapolating backward from modern species. Researchers have extracted DNA from specimens spanning 700,000 years—from ancient horse bones in permafrost—directly documenting genetic changes across hundreds of thousands of generations. This temporal dimension reveals evolutionary processes in action, showing which genes changed during species formation, domestication, or adaptation to new environments.

The technology has rewritten human prehistory by uncovering ghost populations that left no archaeological trace except their genes. Ancient DNA revealed that Denisovans, a previously unknown human lineage, interbred with our ancestors in Asia, contributing genes that help modern Tibetans survive at high altitude and Pacific Islanders resist certain infections. By sequencing ancient humans from across Europe, researchers discovered that modern Europeans descend from at least three distinct ancient populations that mixed at different times, overturning simplistic migration models.

Beyond human history, ancient DNA illuminates extinction events and climate adaptation across species. Woolly mammoth genomes reveal the genetic decay that preceded their extinction as populations shrank to small, inbred island groups. Cave bear DNA shows population crashes corresponding to ice age climate shifts. Crop and livestock genomes from archaeological sites document which genetic variants were favored during domestication, identifying the specific genes that transformed wild wolves into dogs or wild grasses into wheat—changes that shaped human civilization itself.