J-147
guinea pig
The Dose Guy’s verdictSep 1, 20265 sources
J-147 is the rare preclinical compound that moved into real trials. A completed phase 1, a phase 2 recruiting for stroke, and animal data that holds up across multiple aging and Alzheimer’s models. No human efficacy results have been published, and the community running it now is ahead of the evidence. I’d wait for the phase 2 readout. If you want a nootropic you can run now, semax and noopept have human dosing behind them. If you insist on J-147, at least the starting point is cleaner than most compounds on this shelf.
What it is
J-147 came out of a phenotypic screen, meaning researchers tested thousands of compounds against aged neurons in a dish and picked the ones that kept them alive, then optimized from there. The starting chemistry was curcumin-inspired, but the final molecule shares almost nothing with curcumin and works through a completely different target: ATP synthase, specifically the alpha subunit ATP5A, the engine that makes cellular energy in mitochondria. Hitting that target triggers a calcium and AMPK signaling cascade that, in animal models, preserved mitochondrial health, boosted neurotrophic factors like BDNF and NGF, and rescued cognition in old and damaged brains.
On the nootropic shelf, J-147 sits in the preclinical tier alongside ISRIB, but it has moved further: an actual phase 1 is complete and a phase 2 is enrolling. That pipeline is the difference between interesting and potentially real.
What the research shows
The animal data is extensive and consistent. The human pipeline is real but has not published results:
Realistic protocols
I’d wait for the phase 2 readout before running J-147 myself. The community is not waiting. People take 5 to 30 mg orally, typically once daily with food, for weeks to a few months before reassessing. That dose range is a rough scaling guess with no published human confirmation behind it. The completed phase 1 tested single ascending doses, but those results are not in the literature. If you run it now, keep the dose low and track what you notice. You are currently the only safety dataset that matters to you.
Side effects
The cell-level safety data is cleaner than most on this shelf, but no human safety data has been published:
