Open-Source Alzheimer's Research

Scientists Reversed Alzheimer's in Mice. We're Building the Bridge.

The gap between a mouse study and a human therapy isn't just biology — it's tooling. We build the open-source AI infrastructure to close that gap. Not-for-profit. Radically open.

"My mother died from Alzheimer's. My sister was diagnosed 18 months ago. This isn't a research interest — it's a race against the clock I can hear ticking in every phone call home."

David Scott Lewis

Founder, The Reversal Initiative

55M
People living with dementia worldwide
0
Open AI tools for Alzheimer's drug design
10+
Druggable targets in the NAD+ pathway
37+
Pipeline modules built and validated

Three Pillars of Open Drug Discovery

Every tool we build is open-source, peer-reviewed, and designed so any lab on Earth can use it — not just those with billion-dollar budgets.

SCREEN Targets

AI-driven target identification across the NAD+ biosynthesis pathway. Our models evaluate binding affinity, selectivity, and blood-brain barrier penetration to prioritize the most promising intervention points.

BUILD Molecules

Generative chemistry pipelines that design novel neuroprotective compounds. Each candidate is optimized for drug-likeness, synthesizability, and predicted efficacy against validated Alzheimer's mechanisms.

VERIFY Safety

Multi-scale toxicity prediction before any compound reaches a living cell. Off-target binding, hepatotoxicity, cardiotoxicity, and mutagenicity are all flagged computationally — reducing animal testing and accelerating timelines.

Peer-Reviewed Breakthrough

P7C3-A20 Reverses Alzheimer's Pathology in Mice

Published in Cell Reports Medicine (2025), researchers demonstrated that the P7C3-A20 compound — an NAD+ pathway activator — reversed cognitive decline, reduced amyloid plaques, and restored synaptic function in Alzheimer's mouse models. This is the foundational biology our AI tools are built to accelerate.

DOI: 10.1016/j.xcrm.2025.102535

What Our Research Has Built

Every tool we build is open-source. We publish results and cost breakdowns quarterly.

23+
Peer-Reviewed Papers
11
Conference Venues
10
AD Targets Validated
37+
Platform Modules

Where This Work Stands

AAAI-26 Singapore

Featured at AAAI-26 Singapore — 3 oral presentations

NeurIPS Workshop

AI for Drug Discovery workshop with peer review

Stanford Collab

Methodology validated by Stanford computational biology group

2 Drug Programs

Tools actively used in commercial drug discovery pipelines

Someone is diagnosed every 3 seconds

Every Tool We Build Is Open-Source and Outlasts Any Single Trial

Drugs fail. Tools compound. Our research doesn't target one experiment — it builds infrastructure that every Alzheimer's researcher on Earth can use, forever.

Open-source · Free for researchers · Not-for-profit