A token-powered research engine for drug repurposing.

DeCure connects AI search, molecular compute, wet-lab validation and open publication. $DECURE coordinates the work.

A computational result can earn attention. Only experimental evidence can earn confidence.

Fixed supply of 1 trillion $DECURE, minted once: 6% launches liquidity, 94% is the locked protocol reserve.
No presale, no private round, no team allocation.
Future releases split 50% research vaults · 30% Proof-of-Folding · 20% staking.
Proof-of-Folding, staking and research routing are later modules. Each ships with its own public status and official address.
Official mint, pool and program addresses are published on the Addresses page when they are live.

The protocol

DeCure searches approved medicines for new uses. AI maps the biology, structured evidence ranks the candidates, Proof-of-Folding models the strongest drug–target pairs on community hardware, a real laboratory tests what survives, and the evidence is published open-access.

5,000

diseases mapped across 12 research labs

fixed supply — 6% launch liquidity, 94% locked reserve

50 / 30 / 20

each release: research vaults · folding work · staking

01The problem

A promising new use for an existing medicine can be scientifically worth testing and still have no obvious path to funding. The molecule may be generic, second-use protection may be narrow, the disease may be rare, or the first experiment may be too early for commercial capital and too applied for a conventional academic grant. The bottleneck is often not the science. It is who pays for the experiment that settles the question.

02The idea

DeCure is built for that gap. AI maps disease biology and proposes candidates across thousands of diseases, structured evidence ranks them independently of the language model, community compute folds the strongest drug–target pairs, and a qualified laboratory runs the assay that can falsify the hypothesis. Methods, raw data and analysis are published either way. A well-executed negative result closes a path and improves the next decision.

03Proof-of-Folding

Proof-of-Folding is a proof-of-useful-work mechanism for molecular screening. Instead of burning energy on arbitrary hashes, participants run molecular co-folding (a Boltz-class model) to model how a candidate drug binds a disease target. A validator independently re-folds selected work and attests the result on-chain; verified, novel work earns a share of the epoch’s release. A strong fold is evidence about whether the next experiment is worth running. It is not the experiment itself.

Discovery

AI scientists propose approved-drug → disease candidates

Folding (PoUW)

community compute scores binding; validator re-folds & attests

Wet-lab

top candidates tested to prove the effect (~$50K/cure)

Open paper

results published open-access as a citable preprint

04The 12 labs

Research is organized into 12 autonomous AI-scientist labs — one per disease area — collectively covering 5,000 real diseases, many with no approved therapy today (the "white space" where the need is greatest).

Cancer · Neuro · Psychiatry · Cardio · Respiratory · Nephrology · Dermatology · Metabolic · Immuno · AMR · Longevity · Rare & Orphan.

05$DECURE — the token

$DECURE does three jobs: it rewards verified molecular compute, it lets stakers direct research across the twelve disease areas, and it builds the budgets that pay for experiments. Fixed supply, minted once, with no presale and no team allocation. It carries no equity, dividend, royalty or claim on a paper, patent or laboratory, and it can lose all value.

Property	Design
Launch	Fair launch on Solana — no presale, no insiders
Allocation	6% launch liquidity · 94% locked protocol reserve
Emission	50% to the research vaults · 30% to the folding work pool · 20% to staking
Staking	One position, one research area; longer locks carry more weight
Rewards	Variable, protocol-defined, never guaranteed

Staking tiers

Tier	Lock	Boost
Bronze	7 days	+5%
Silver	30 days	+12%
Gold	365 days	+25%

How staking directs the science

One position, one lab — every staking position points at one of the twelve research areas and adds weight to its share of future research releases.
Bounded outcomes — each lab receives at least 5% and at most 40% of a research allocation epoch, so attention can move without any single field taking everything.
Evidence decides — stakers choose where attention goes. Whether a result is true is settled in a laboratory, not by a vote.

06Governance

A light, on-chain DAO — scope limited to steering the protocol, not a legal entity. Stake-weighted voting sets the allocation across the 12 labs, each bounded to a 5–40% share. Longer locks carry more weight, and the 5% floor plus the 40% cap keep any single holder from buying the outcome.

07The mission

The ambition is deliberately large: to channel up to $1 trillion into real drug-repurposing research, published in the open, and to make curing a neglected disease something a global community can fund and direct together. "Trillion-dollar" is the size of the mission — not a price target, forecast, or promise about the token.

Risks

The honest version.

$DeCure may lose its value in part or in full, and there is no mechanism and no party obliged to prevent that.
It may not always be transferable, and it may not be liquid.
It may never be exchangeable for anything, in particular if the project is discontinued.
It is covered by no investor compensation scheme and no deposit guarantee scheme, because it is not a financial instrument or a deposit.
Nobody is responsible for making the project succeed and nobody can be compelled to continue it.

Scientific risk

Drug repurposing has a high failure rate. Most candidates will not become therapies.
The prioritisation engine's measured discriminative signal is weak: held-out AUROC 0.60 (95% CI 0.578–0.622; n = 2,708 pairs, 455 diseases; label-shuffle control 0.50). It is published with its interval and shuffle control precisely so it is not mistaken for predictive certainty.
Prior human exposure to an approved drug removes no requirement for dose, route, population, combination or new preclinical and clinical work in a new indication.
Repurposed generics carry weak or fragmented exclusivity — the reason the field is under-financed, and a constraint on the protocol's own sustainability.
Peer review is outside anyone's control and cannot be guaranteed on any timeline.

Technical and structural risk

Nothing is deployed. There is no assurance the protocol will be built, deployed or work as described.
Unaudited code, key management failure, irreversible transactions.
Dependence on a single third-party blockchain, including outages, congestion, forks and validator concentration.
Emission concentration: 94% of supply sits in the locked reserve, which materially affects circulating supply and price formation for as long as it is being released.
Attack on the useful-work mechanism itself — participants attempting to claim rewards for work that is not novel, not correct, or not actually performed. Validator re-folding and attestation exist to counter this and may prove insufficient.
Regulatory risk: the treatment of the asset in any jurisdiction may change, and the position set out above may be rejected.

Current state

Component	Status
Disease catalogue	Live — 5,000 pages across the 12 labs
Evidence pipeline	Operating — retrieval over 234,678,978 indexed works, experimental structure resolution, relationship-direction audit
Verified drug–disease pairs	397 pages, 525 pairs, corroborated in primary literature and audited for relationship direction
Prioritisation engine	Held-out AUROC 0.60 (95% CI 0.578–0.622)
Proof-of-Folding program	Coming soon
Token	Coming soon — no mint address exists
Network	Coming soon
Wet-lab validation	Coming soon
Peer-reviewed publications	Coming soon

Disease pages and model outputs are research, not medical advice, and no claim is made that a cure exists for any disease. $DECURE confers no equity, dividend, royalty or claim on a paper, patent, dataset or laboratory. Most candidates fail in the lab. Research outcomes are never guaranteed and $DECURE can lose all of its value.

Protocol paper v1 · Token stack · Disease catalogue · Project