The double helix is a toy reference genome. Short fragments ("reads") are extracted from the ancient sample and drift toward it; wherever a read lands, that stretch of genome lights up as covered. Two real paleogenomic signatures drive the fragments:
L(t) = L₀ · e^(−t/τL) avg. surviving fragment length
D(t) = Dmax · (1 − e^(−t/τD)) terminal C→T deamination rate
- Sample age (t) — older bones/teeth yield shorter, more heavily damaged DNA (both curves above depend on it).
- Sequencing depth — how many reads/second are generated; higher depth fills genome coverage faster.
- Modern contamination — a fraction of reads are undamaged present-day DNA (grey), diluting the authentic damage signal — exactly the problem real paleogeneticists must screen for.
- Damage signal — the measured share of genuine ancient reads showing the 5' C→T misincorporation. Real ancient DNA almost always shows this pattern; contamination doesn't, which is how labs authenticate a sample as truly ancient (e.g. Neanderthal or Denisovan) rather than modern handling contamination.