5 fields where re-reading history reveals something new · built to be audited
Every one of these turns data we already collected into knowledge we couldn't extract at the time. If verified, the real-world payoff is the same in each case — we don't always need to gather more; we need to read what we have more carefully.
Honest risk: Re-analysis is interpretation. Every finding is a hypothesis from old data until fresh measurement confirms it.
Analog method
Single-indication clinical trials — one drug, one endpoint, approved and shelved for everything else.
Modern re-analysis
Network pharmacology + multi-omics + causal inference over decades of post-market adverse-event reports (FDA FAERS) and electronic health records.
What it reveals
Existing drugs have off-target effects invisible to single-endpoint trials. Modern ML over the full adverse-event graph surfaces new indication candidates that analog 'one drug, one disease' approval logic buried.
Real, documented
Sildenafil (angina → erectile dysfunction), thalidomide (nausea → multiple myeloma), metformin (diabetes → cancer / aging signal).
If verified — layman implication
A drug already proven safe — maybe in your medicine cabinet — turns out to treat a second disease it was never designed for. Faster, cheaper paths to new treatments: ~10 years and billions in trials skipped because the safety profile is already known.
Caveat: Must be verified in new trials — a database signal is a lead, not a cure.
REAL: the named cases (sildenafil, thalidomide, metformin; 1918 cytokine-storm reframing; DASCH plate-digitization; Gray & Atkinson 2003; Steppe-ancestry aDNA; Caracol/Angkor lidar) are documented in the scientific and historical record. MODELED: the layman 'implications' are reasoned extrapolations of what each finding means for everyday people, not measured outcomes. NOT CLAIMED: that any re-analysis alone constitutes discovery — every finding here is a hypothesis from old data until fresh measurement confirms it. This page declares the fields, the methods, and the verification gap honestly rather than promising the implications as guaranteed.
Analog → Digital Implications Lab · read what you already collected, more carefully.