Findings cryptographically timestamped · March 13, 2026
Three drug repurposing candidates for Alzheimer's disease and three clinical biomarker targets were identified in a single computational session using a novel semantic compression framework. The framework encodes pharmacological knowledge as causal graph topology rather than categorical labels, enabling traversal across domain boundaries that conventional databases cannot cross. The principal finding identifies a dual-mechanism pathway by which Metformin intersects the Alzheimer's neurodegeneration axis. All findings are cryptographically timestamped and reproducible.
Two compounds converge at the same neuroprotective protein through completely different entry mechanisms. Neither is categorically adjacent to Alzheimer's treatment in any existing database.
Five drugs from five unrelated pharmacological categories converge at the same protein node through five distinct mechanisms. No existing database has ever placed these five drugs in the same sentence.
A sixty-year-old generic diabetes medication intersects the Alzheimer's neurodegeneration axis through two simultaneous mechanisms. This provides a mechanistic explanation for decades of unexplained epidemiological observations.
The framework resolved findings to three specific biomarkers used in clinical Alzheimer's diagnostics: p-tau181, Aβ42, and IL-1β. The same markers your neurologist orders in blood panels and PET scans.
The drug already exists. The safety data already exists. What was missing was the mechanism. We found it by reading what we already knew — differently. This is the first disease we tackled. It will not be the last.
Type a compound. Watch the framework traverse across domain boundaries in real time.
The INZ framework encodes scientific knowledge as causal graph topology using a proprietary formal compression grammar. The framework operates on sovereign local hardware. Each discovery event is cryptographically timestamped at the moment of generation.