Sofia pulled up the raw data again. There it was — the same secondary structure shift predicted by her mfold simulation. A G-to-A transition at position 7,421 of the long non-coding RNA NEAT1 . The change didn’t touch any splice site or coding exon, but it created a new hairpin loop that sequestered a microRNA critical for neuronal survival.

Tonight, that sentence felt prophetic.

She was studying a rare genetic variant found in the DNA of a young boy from Sardinia — a mutation in a non-coding RNA gene that no one had bothered to characterize. Everyone had told her it was “junk.” But page 164, in a footnote she’d almost missed, described a similar case from 2003: a silent mutation that altered RNA folding, not protein sequence. The boy in that paper had developed late-onset neurodegeneration.

Her hands trembled. She had just found the mechanism.