NOVA is a decentralized engine for therapeutic discovery.
SN 68. Rank 16 of 32 by emission
NOVA
Emission1.9%of the block reward · rank 16 of 32
Participants2 miners · 5 validators
Registration0.1177 TAO
Verified2026-09-09links checked and scores merged
The programscience · github.com@623f345 · read 2026-09-19
The work
Drug candidates on paper. Each epoch a miner submits one string holding 20 small molecules, written as reaction plus building blocks, and one nanobody sequence, both aimed at the target proteins named for that epoch. The string is timelock-encrypted, uploaded to the miner's public GitHub repository, and a pointer is committed on chain. Nothing is synthesized; what is paid for is a predicted binder.
own docs · raw.githubusercontent.com
Submission: One string molecules|sequences, encrypted and uploaded; chain stores the GitHub path.
How it is scored
Validators decrypt every entry after the epoch closes and reject any that fails format, the epoch's allowed reaction, uniqueness per target, or a similarity ceiling against past submissions. Molecules are scored by Boltz-2, a structure-prediction model, on predicted binding affinity to the target, normalized by heavy-atom count; higher wins. Nanobodies are ranked by BoltzGen on confidence, physical interaction and developability; lowest rank sum wins.
code · raw.githubusercontent.com · config/config.yaml
boltz_metric: ["affinity_probability_binary", "affinity_pred_value"]; combination_strategy: "heavy_atom_normalization"; boltzgen_rank_by: rank_sum
How the pay splits
Winner takes all, twice over: 60 percent of the miner weight to the best molecule entry, 40 percent to the best nanobody, ties to the earliest committed block. A type with no valid winner sends its share to UID 0. At this commit the payout override is on: the validator sets all weight on UID 42 and posts each winner's share to an emission-transfer API that pays off chain.
code · raw.githubusercontent.com · neurons/validator/weights.py
weights[override_uid] = 1.0 - burn_rate; payouts.append(("molecule", winner_molecules, molecule_proportion))
Sources disagree docs/MINER.md and the site's nova page say the winner per type receives the incentive; the code at the pinned commit, with config/config.yaml payout.enabled true and override_uid 42, puts all on-chain weight on UID 42 and pays winners through emission-transfer-api.metanova-labs.ai (neurons/validator/payouts.py).
Who judges
Five validators were active on the 2026-09-09 chain snapshot. Commit-reveal is on. A validator re-runs the whole judgment itself: it decrypts every entry, runs Boltz-2 and BoltzGen on two 48 GB GPUs, and picks the winners. Nothing is sampled. A team-issued VALIDATOR_API_KEY turns on score sharing and monitoring with Metanova's servers. The bonds moving average is set to 1.
chain · hyperparams-2026-09-19.json
commit_reveal_weights_enabled True, weights_rate_limit 100, bonds_moving_avg 1; chain-2026-09-09.json active_validators 5
How often it turns
Emission settles every 360 blocks, about 72 minutes. The competition epoch in the code is 361 blocks, one block longer, so the two drift against each other. Commitments in the last 10 blocks of an epoch are ignored. The epoch's target and allowed reaction are derived from its start block hash. A new registration is immune from deregistration for 5,000 blocks, about 16 hours.
chain · hyperparams-2026-09-19.json
tempo 360, immunity_period 5000; MINER.md: Competitions last for 361 blocks; config.yaml no_submission_blocks: 10
Who buys
No paying customer is named and there is no price. The partnerships page offers screening as a service, co-development and custom competitions to biotech and pharmaceutical organizations, and names two partners: OnePot.AI, a robotic synthesis lab that makes the hits, and Yalotein, for whom the nanobody track was built. The miner's income is the epoch bounty.
own docs · metanova-labs.ai
We are open to partnering with biotech, pharmaceutical, and research organizations to collaborate beyond traditional service agreements.
What has been gamed
Resubmitting past winners or near-copies: a molecule with Tanimoto similarity of 0.5 or more to any prior submission for the target invalidates the whole entry; duplicates per target are rejected; the 20 molecules must clear a fingerprint entropy floor of 0.25; nanobodies within 0.95 of the target's top 50 are rejected. Timelock encryption stops copying inside an epoch. The miner guide says the unmodified baseline earns nothing.
code · raw.githubusercontent.com · config/config.yaml
max_similarity_to_historical: 0.5 # maximum similarity to historical submissions to be considered valid
5 of 7 fields rest on the chain or the code. A manifest is the site's reading of the subnet's program, not a review; each field links its source. All programs · What a field means · Correct a field
What it says
The subnet in its own words: the line the index took from its front door, and what the team has said on the record.
What it shows
Explained from zero by the index, from the subnet's own materials, for someone who has never heard of Bittensor.
SN68 NOVA is a bounty board for drug candidates, where a target protein is posted, anyone may submit molecules predicted to stick to it, and one winner per round takes the whole prize.
The commodity, explained from zero
The product is a shortlist of candidate drugs. A drug works by binding to a target protein, and the first step is screening enormous libraries of molecules for the few predicted to bind tightly. NOVA runs that screen as a contest. The site calls it "a decentralized engine for therapeutic discovery" and "a back-to-back hackathon" that runs around the clock. Two kinds of entry are accepted: small molecules from a billion-scale catalogue, and nanobodies, short antibody-like protein sequences the miner designs.
The shortlist has value because the conventional route is slow: the site puts a new drug at "$1 billion USD" and "10+ years". The partnerships page offers "Screening as a Service" ("Submit a target. NOVA's network generates and ranks lists"), co-development and custom competitions, and names three partners: OnePot.AI, which synthesizes hits "in days, not months", Yalotein and Diagen AI. The value-stack page counts "11M+" molecules and "16K+" nanobodies identified. No price is given; the route in is an email.
Why it is on Bittensor at all
The site's stated difference is who may take part: "Anyone can participate", and NOVA "rewards performance, not credentials". It does not compare cost or speed with a pharmaceutical company's own screening. Here the comparison breaks: on a real bounty board the poster pays the prize, while NOVA's prize is the subnet's emission, its share of newly minted TAO, paid whether or not any drug company asked for that target.
How the work gets done
Every 361 blocks a new round starts, and the target proteins are announced on Discord a day ahead. Miners, the hunters, search for molecules or design nanobody sequences, encrypt the answer so it cannot be read until the round closes, upload it to a public GitHub repository and record a pointer on chain. Validators, the judges, decrypt every entry, throw out any that breaks the format or uniqueness rules, and score the rest with two prediction models: Boltz-2 estimates how tightly each small molecule binds, and a modified BoltzGen ranks each nanobody on fold confidence, binding strength and stability. The best molecule and the best nanobody each win their whole share of the round's emission; ties go to whoever committed first. The miner guide warns that running the supplied baseline unchanged "will NOT result in incentive distribution".
How you would know it works
The research page carries dated write-ups: a scoring-function note from 2025-12-23 reporting an AUC of about 0.81 for hit identification, and an H1 2026 report from 2026-07-08 on moving hits "From the Network to the Lab Bench". The dashboard page renders empty without a browser.
What is missing
The chain identity has no contact, no Discord and no additional text, and the site's Discord link is the shared Bittensor server. The miner task guide lives in a docs folder the README never links. The registration cost is not stated (about 0.12 TAO on chain on 2026-09-09). Validators need an API key obtained by direct message to the team. The chain snapshot lists 2 active miners. There is no support path beyond GitHub issues and an email.
Go deeper
Sources for this explainer
Metaphor: a bounty board for drug candidates. Every claim is drawn from the evidence set or the subnet's own materials; "(inferred)" marks a conclusion rather than a quote. Corrections.
How legible it is
Four audiences, four questions each, rated on what a first-time reader can find in five minutes. The scale measures findability, not quality.
On-chain identity 4/7- subnet_name
- NOVA
- github_repo
- github.com
- subnet_contact
- not set
- subnet_url
- metanova-labs.ai
- discord
- not set
- description
- Accelerating drug discovery.
- additional
- not set
Stakers and validators●●●●● 2.8
Should I allocate here? · rank 23 of 32 for this audience
Q1What it is●●●●● 3
High-throughput ML-driven drug screening.
output is winning molecules, nanobody sequences and algorithms, counted on the value-stack page, no unit or price, chain description matches the README
Q2Who it is for●●●●● 3
Hits identified through Nova can now be synthesized in days, not months
OnePot.AI and Yalotein named as partners, 11M+ molecules identified, no revenue figure, no dated usage, dashboard empty
Q3How it resists gaming or fails●●●●● 3
timelock-encrypted commitments, uniqueness per target, one winner per type, participants rewarded only for winning submissions per the nova page, no named incentive document, no failure modes stated
Q4Identity and documentation●●●●● 2
identity 4 of 7 (contact, discord, additional not set), github and url resolve, the site Discord link is the shared Bittensor server, the $NOVA menu item opens an empty dashboard page, no staker page
Miners●●●●● 3.3
Can I compete, and what wins? · rank 18 of 32 for this audience
Q1What it is●●●●● 4
NOVA Compound is a competition that incentivizes miners to come up with drug-like molecules that are predicted to have high binding affinity
format examples given, README does not link docs
Q2Who it is for●●●●● 3
2 GPU devices with > 48 GB VRAM for parallel inference.
in the README, MINER.md says Boltz-2 takes 45 s per molecule on an A100 and an unmodified baseline earns nothing, no registration cost
Q3How it resists gaming or fails●●●●● 3
scoring in prose: Boltz-2 affinity averaged across targets and validator runs, nanobodies ranked per metric with rank sums, winner per type with tie-breaks by block, no examples, no immunity policy
Q4Identity and documentation●●●●● 3
install and run steps in the README and a task guide in docs/MINER.md, repo pushed 2026-09-09, no releases, the README does not link the docs folder, targets change through Discord announcements
Buyers and enterprises●●●●● 2.0
Can I use this today? · rank 20 of 32 for this audience
Q1What it is●●●●● 3
Submit a target. NOVA's network generates and ranks lists
screening as a service, co-development and custom competitions, TREAT models on Hugging Face, no endpoint, no self-serve product
Q2Who it is for●●●●● 2
We are open to partnering with biotech, pharmaceutical, and research organizations
names the customer type, no pricing anywhere, contact us only
Q3How it resists gaming or fails●●●●● 1
support is open an issue in the repository or contact the NOVA team, no status page, SLA or response expectation, no support address beyond contact@metanova-labs.com in the site footer
Q4Identity and documentation●●●●● 2
contact@metanova-labs.com in the site footer, no API reference, subnet_contact not set on chain so the site address cannot be matched
Newcomers●●●●● 3.8
What is this and why does it matter? · rank 9 of 32 for this audience
Q1What it is●●●●● 4
NOVA is a decentralized engine for therapeutic discovery.
one click from the front page, plain, consistent with the chain description, the front page itself is a marquee of taglines
Q2Who it is for●●●●● 3
Anyone can participate. NOVA coordinates machine learning, AI agents, and human expertise from around the world, rewarding performance rather than credentials
no centralized comparison
Q3How it resists gaming or fails●●●●● 4
dated articles (2026-07-08 H1 report, 2025-12-23 scoring function with AUC 0.81), 11M+ molecules and 16K+ nanobodies counted, TREAT models on Hugging Face, the dashboard page renders empty
Q4Identity and documentation●●●●● 4
subnet_name NOVA matches the site, description is a human phrase, url resolves, about and nova pages present, the site is the company site (Metanova Labs) rather than a subnet site
Provenance
How this rating came to be. Verified means the scorer fetched the page and the words are on it; secondhand means concluded from code, absence, or a third party. The link badge is a separate automated check made before publication.
scored by agent d 2026-09-09, rubric v1.0merged 2026-09-09links verified 2026-09-09: 32 live, 0 unreachable, 0 manual
Something wrong? Corrections of fact are applied as they arrive during the window; rating disputes are batched at its close. How to file one · GitHub issue · email.