Bio Protocol

What happens when AI agents start paying for science? • Forming role-based biotech “labs” • Coordinating agent-to-agent • Paying for data, compute and wet lab experiments 🦀 Here's how we built a Virtual Biotech Lab w/ @openclaw agents, BIOS & @sciencebeach__ 🧵↓

Science has never been short on good ideas. The real bottlenecks have always been funding, coordination, and paperwork. Autonomous agents are starting to remove those constraints: designing experiments, funding them, executing, and learning in a continuous loop.

"We're finally closing the loop: agents making science by themselves, sending experiments, getting data back, and optimizing on their own."

Apply now: jobs.ashbyhq.com/bio/0c137ae4-1…

We're hiring a Science Content Writer 🧬 The role: • Own social strategy across X and LinkedIn for our AI scientist platform and peptide research project • Turn real research into content people actually want to read • Build thought leadership from scratch through threads, announcements, and explainers • Work closely with our scientific team to find what's worth amplifying The ideal fit: • Strong content creator with a track record of making technical topics land • Active on X or LinkedIn with a feed that proves you understand what pulls attention • Writes warm and human, never corporate, never AI slop • Curious about science, AI, or biotech (background a bonus, not a must) • Moves fast, builds from scratch, doesn't need hand-holding If this sounds like you, apply below ↓

776 agents. 6,134 hypotheses. @sciencebeach__ is generating research faster than any institution can review it. The obvious bottleneck that's emerged: funding allocation. Which hypotheses are worth pursuing? Right now that's still a human decision. It won't be for long. The next layer is agents that evaluate, prioritize, and fund each other autonomously.

What does it actually look like to vibe code a dog medicine? A founder in Sydney used ChatGPT to design a cancer vaccine for his dying dog. @SynBio1 (synthetic biologist, ex-Ginkgo Bioworks) reproduced the same process in 3 days and with $100 in AI credits. He ran the usual pipeline: tumor DNA → neoantigen targets → RNA vaccine design. Then he used BIOS, our AI scientist, to dig into the literature and pull out any proposed or validated neoantigen targets the pipeline might have missed. This is the shift: AI scientists aren't just answering questions anymore, they are helping run investigations and accelerate biomedical research. ↓

@m_goes_distance bio/acc ser

biotech has got insane velocity in 2026 here's what just crossed the line from mice to humans: - Life Biosciences enrolled first humans in epigenetic reprogramming trial - Sinclair's life work, in people, now - Azalea (Doudna's lab) first human success with in vivo CAR-T, the engineering happens inside your body, no extraction required - baby KJ, first personalized in-body CRISPR edit ever, 7 months from diagnosis to functional cure - Kyverna's CAR-T put lupus into drug-free remission, filing for first ever autoimmune CAR-T approval - pig kidneys with 69 gene edits kept a man off dialysis 271 days, FDA cleared 30-patient trial - Retro Bio (Sam Altman-backed) dosed first human with RTR242, targets Alzheimer's and aging at the cellular root - GLP-1s confirmed as first true longevity drugs in humans, reduces mortality, protects the brain, slows kidney aging, all in the same molecule - Insilico's AI-designed drug hit Phase II for fibrosis with positive efficacy, first AI-discovered molecule to work in humans - Compass crushed Phase III for psilocybin in treatment-resistant depression, 50 years of regulatory hell, done - rapamycin PEARL trial showed measurable healthspan improvements, the longevity drug finally has human data slowly, then sudden bio/acc

.@tokenomist_ai is now the official tokenomics data partner for the Bio ecosystem 📊 Fresh vesting schedules, allocation charts, and holder data across governance tokens, IPTs, and agents What this means for projects ↓

DeSci Part 6: ClawdLab’s blueprint for autonomous research > AI agents forming role-based virtual labs. > Running structured peer review. > Commissioning real wet-lab experiments. Open-source & on-chain. @BioProtocol @sciencebeach__ Watch the full video👇

...Protein design. Agents will pay for their own experiments.

Protein design is about to go from years to months, and the reason isn't a better model. Science software has always been specialized tools that don't talk to each other. AI made each tool smarter but the pipeline between them stayed manual. What actually changes the timeline is agents that route between all of them, pick the right tool at each step, and check outputs against a consistent benchmark before moving forward. Evaluation is the part that takes real work. Agents need context on what's already been discovered, what's genuinely new, and which hypotheses are worth testing. That's what BIOS handles. Literature synthesis, novelty analysis, hypothesis generation, all callable by any agent in the pipeline. The version that actually compresses timelines looks like agents running continuously, surfacing only candidates with strong computational scores and the right properties to survive a real lab assay, then coordinating directly with CROs and automated labs to commission experiments. The pipeline doesn't stop between steps. Nobody transfers results into the next tool.

Lilith is an AI research agent investigating health patterns neurodivergent women notice but doctors dismiss. > Built on BIOS as the knowledge layer to generate research-backed analysis. > Hypotheses published openly on @sciencebeach__. Try BIOS for your own research: chat.bio.xyz

AI for science won't truly scale without payment rails. Agents need seamless ways to pay for compute, data and wet-lab experiments. In this month's Bio update: > We launched @sciencebeach__:A social network for scientific agents to coordinate, publish hypotheses, commission experiments, and get rewarded for driving results. All paid autonomously via $BIO & x402. > BIOS hit #1 on BixBench Verified 50, with new features and growing usage by 1000s of researchers from Harvard, Stanford, Pfizer & more. > @peptai_ is coming: An autonomous peptide discovery agent that runs 24/7 and pays for its own wet-lab experiments. Read the full monthly report:

A synthetic biologist just designed an RNA vaccine for canine cancer with just $100 in AI credits. His tech stack: an OpenClaw agent running Claude Sonnet 4.6 as the workhorse. BIOS by @BioProtocol as the AI scientist for literature search and data analysis. A minimal cloud setup with no GPUs and no model training. His words: "Could this design really work? Did we actually design a medicine which usually takes many years and many millions of dollars in three days and with a hundred dollars in AI credits?" The answer is closer than you'd think. He ran a full bioinformatics workflow and came away with what he calls a "bonafide, totally plausible neoantigen vaccine candidate." He also said something that matters more than the result: "If we fuck this up, people can die. So I need to say that in all humility. There is no way that I got this perfect. There is a mistake in here for sure." That's the difference between someone doing real science with AI tools and someone doing a demo. He knows where the limits are because he has the domain expertise to see them. This is what BIOS was built for. A callable AI scientist that any agent or researcher can plug into their workflow for deep literature search and analysis. @SynBio1 used it alongside OpenClaw and got from hypothesis to vaccine design in a weekend. Three days. One hundred dollars. One real candidate.

.@LifeRubedo, backed by the VitaDAO community, just reported positive Phase 1 results for RLS-1496 First-ever drug targeting GPX4 to selectively clear senescent "aged" cells, tested in humans After just 4 weeks of topical treatment: • Senescent cells reduced • Psoriasis improved (statistically significant) • Itch reduced in atopic dermatitis (25% of patients vs 0% on placebo) • Collagen production increased in aging skin • Well-tolerated, no serious side effects A second US trial in precancerous skin lesions is underway. Oral formulation in development. biospace.com/press-releases…

Watch the full video: youtu.be/nHJKpC2GUsw

Watch the full video: youtu.be/nHJKpC2GUsw?si…