The sequencing itself — the actual reading of the nucleotide sequences in the DNA fragments — produced hundreds of millions of short sequence reads.

Each read was a short stretch of sequence, perhaps fifty to a hundred nucleotides long.

These reads were computationally mapped to a reference human genome, allowing researchers to determine which parts of the genome each read represented.

The authentic ancient reads were identified using the characteristic damage patterns of ancient DNA: elevated rates of cytosine deamination at the ends of fragments, which produces specific sequence changes that are rare in modern DNA but diagnostic of ancient DNA.

By filtering for reads with this damage signature, researchers could separate the genuine ancient signal from the contaminating modern DNA that inevitably found its way into the sample despite the precautions taken.