For a long time, the roughly 98% of the human genome that doesn't code for protein was dismissed as "junk." That label has aged badly.
What changed
Projects mapping regulatory activity across the genome found that huge stretches of non-coding DNA are doing real work: switching genes on and off, structuring how DNA folds inside the nucleus, and producing regulatory RNA molecules.
Why it matters for our work
A lot of the variants we study in early-onset risk prediction sit in these non-coding regions. Ignoring them the way older models did means missing real signal.
What we still don't know
Most non-coding variants still have no confirmed function. The honest position is that some of the genome probably is inert — we just don't yet know how much.