Bio Protocol

Highlights and key takeaways from DeSci Summit Mumbai this week 👇 → The IP question decides whether a biologist can use frontier models at all. Several PhD labs bar ChatGPT and Claude outright, since prompting them leaks unpublished research. → Ask a frontier model to design a 10 amino acid peptide that binds and inhibits the SARS-CoV main protease and the request gets flagged. Both refusals ran live on stage, and the same prompt at @biosdiscovery returned an answer plus a route into a binder generation pipeline. → Finding a binder is finding a needle in a haystack. Doing it in silico only became possible after @GoogleDeepMind in 2020 and RFdiffusion from the Baker Lab. → Ten scored binders come back within ten minutes. The models sit on 200 GPUs, which opens the platform to anyone who shows up today. → RFdiffusion 4 runs on Mac, ported by @GeeveGeorge and currently the fastest RFdiffusion MLX model. A scientist designing a drug keeps every piece of data on their own machine. → A Qwen 8B model runs at 100 tokens/sec on M1 through M5 using custom kernels, ahead of ZSpec and Splash. The team calls it spec metal. → Qwen orchestrates and RFdiffusion designs, so drug discovery runs fully locally. The @biosdiscovery CLI is agent-native, built on the view that agents will be the ones using this software. → In coding and math the verifier is fast. In biology the verifier is the world, which is why wet lab demand rises as more people design medicines with AI. → Robios ran its first live demonstration. It loads and unloads the machine and records conditions like temperature, so two runs can be compared. → Human in the loop throughout. A senior scientist watches from home, gives feedback at any time, and steers it mid-run. → Competitors buy their robots from OEMs at prices most labs are unable to afford. Robios is built on open source and aimed at NCBS Bangalore and a wet lab in Mumbai. → Local agentic swarms are next, 10 to 20 agents on a MacBook reading real research papers and citing them. Citation is how hallucination gets reduced. → A research proposal takes around 116 hours to write. Grant acceptance has declined to 19% in the UK, 13% for US health research, and 8% for EU early career awards. → Publication costs roughly $12,000 at Nature and 17 months. @openlabsbio puts those first steps in one place, with an index covering failed and ongoing work, challenges for calling in skills, vaults funding first studies, and a launchpad for pre-company raises. Idea, lab, funded IP, product. Every layer we build shortens one of those and opens the next. Full talks ↓

Robios in action at DeSci Mumbai! AI-designed molecules still need to be tested in the lab. With Robios, we’re exploring how robotics could help researchers turn those designs into experiments and use the results to guide what comes next.

From an initial question to a project developed in public, Ryan is showing how researchers can use @openlabsbio to find tools and work alongside AI agents and other contributors to move their ideas forward.

Livestream on YT: youtube.com/watch?v=FeNNpJxT…

Tuesday Motivation: Building a system where each generation of research helps fund the next.

Patients should have a say in which research questions get pursued. That means being involved early, when priorities are set and projects take shape with a clear view of how decisions are made. Science should belong to more people.

Decentralized science is coming to Mumbai 🇮🇳 The Bio community is bringing researchers, founders and builders together for an afternoon of DeSci + AI with the first live demo of our lab robot, Robios. Join us November 5 on the same grounds as @EFDevcon. Sign up to attend and meet IRL ↓

Before you build on a research tool, follow its source. Bio’s OpenLabs Index brings scientific software, models and datasets into one directory, with licence and activity signals to help you investigate.

Research ideas deserve a place to develop before they’re ready for a paper. OpenLabs (@openlabsbio) gives researchers a place to share a hypothesis and document what they find as they investigate it. Explore what others are working on, or bring your own question.

The $NDT token is officially live on @base. Claim NDT on the Bio App: app.bio.xyz/agents/neurolab-… CA: 0xd89e13e5faee17a8b4dd0739e0a9cab12b3828e3 → NDT rewards are now live and claimable on the Bio app. → veBIO holders receive 20% of the total supply, 6-month linear vesting with 20% redeemable at TGE. Claiming early forfeits any unvested tokens. Check out the NDT roadmap below

An early experiment should clarify a decision: continue, change course, or stop. A negative result can save months of work. The hard part is funding that test before there’s a success story. Bio helps research teams raise community backing for the next milestone.

Finding the right open source scientific software shouldn't be the hard part of your research. Open Science Index groups tools by scientific application. You start from your research problem and compare the options against each other. We check the source registries daily, so you can see what is new and what is still maintained ↓

The next Ignition Sale is live on Bio! @NootropicsDAO_ is building NeuroLab AI to predict how compounds could affect the brain, connecting molecular interactions to potential effects on cognition. The goal: help researchers choose which compounds to test next. Secure your NDT allocation on Bio: app.bio.xyz/agents/neurolab-…

The human brain contains an estimated 86 billion neurons. Here is a glimpse inside a single human neuron. One cell, with an entire world working to keep it alive and connected.

The next Ignition Sale is expected to go live on Bio next week! @NootropicsDAO_ is building NeuroLab AI to explore how compounds could affect the brain, from cognition to potential side effects, and help researchers decide what to test in the lab. Use BioXP + USDC to join the sale on @base ↓

The new Bio website is live. Bio is evolving from a high-touch launchpad into a permissionless factory for decentralized biotech. Our new website brings that direction into focus, with all our products in one place. • Explore research with BIOS and OpenLabs. • Get funding-ready with Buildspace and raise through Launchpad. • Bring research-backed products to your community with Biofy. The path to funding starts where the research happens. Each part feeds the next, with usage data and revenue flowing back into research. Science for the Network Age. Explore the new bio.xyz

Introducing Open Science Index What if every open source science tool, AI model, dataset, and workflow across biology, chemistry, physics, materials science, and earth science lived together in one searchable place? That's the idea behind Open Science Index. Scientific tooling is exploding, and keeping up with what exists, and what still works, is hard. What started as our own internal tracker is now open to every researcher. → Open source tools, models, and datasets across every science domain → Health signals: citations, downloads, commits, maintenance status → Linked papers, benchmarks, and alternatives → Open community curation 6,720 resources indexed so far, built on top of bio .tools, Bioconductor, Bioregistry, Hugging Face, BIOS ( @bioaidevs), various GitHub awesome lists, and across other scientific domains. Each linked back to its source. An open, cross-domain index where any researcher can find tools that fit their work, see what is actually used and maintained, and contribute back as their own tools and reviews help shape the catalog. Try Open Science Index and read our blog post in the links below ↓

Launchpad wallets can now be funded by credit card. Previously, funding a wallet meant buying elsewhere and transferring in. Card funding happens on the Launchpad itself, through @privy_io, and self-custody is unchanged: the wallet stays yours, and Bio never holds the funds. This is the first step toward something larger. Most people who want to fund science don't hold crypto and shouldn't have to. We're building the rails that let fiat capital reach research directly, and removing each piece of friction between wanting to back a project and doing it.

Science pays for the same failure many times over, and nobody involved is doing anything wrong. A lab spends two years on an approach and it doesn't work. That result closes a door, and there are only so many doors, which makes it worth knowing. It goes nowhere. No journal wants it. No grant renewal rewards it. There's a quiet professional cost to saying out loud that two years produced a no. So it stays in the lab, in a drawer, in the memory of a postdoc who has since left for industry. Then another lab has the same idea. It's a good idea, which is why two groups had it. They spend their two years. Same result, same silence. Somewhere a third group is writing the grant application right now. Failure is how the search works. The problem is that each failure gets paid for repeatedly, by people who had no way of knowing it had already been bought. This follows from how research gets funded. The reliable way to be paid is to produce results that look like forward motion, and a no doesn't look like that from the outside. Change who's paying and the pressure changes shape. A group that pooled money to find out whether something works has no stake in the answer being yes. They wanted to know. If it's a no, that's what they bought, and there's nothing to protect by keeping quiet. Most of what follows from that hasn't been built yet. It's a different starting position, and starting positions compound.