Poreva Open Tests 10,000 AI Protein Designs in One Wet Lab
MegaRobo Technologies has launched Poreva Open, a global AI protein design competition in which more than 10,000 designs will go through standardized wet-lab experiments. The company says the aim is to build a shared, experimentally grounded benchmark and move beyond computational predictions.
- Poreva Open is a global AI protein design competition launched by MegaRobo Technologies, with more than 10,000 designs to be tested in the wet lab.
- Five challenges are planned; the first three target EGFR, TNF-α and GLP-1R, and two more come in the second round.
- Prizes total US$250,000 in cash plus US$150,000 in wet-lab validation credits, with up to US$1.5 million in potential investment for the top 10.
- Results and data, including unsuccessful designs, are due to be published in April 2027.
A competition built around lab results
MegaRobo Technologies, which describes itself as a company building "Intelligence Infrastructure for the Physical World", announced Poreva Open as a way to check AI-generated protein designs against real experiments. According to the company, a promising design is only the beginning, and "the real test comes in the lab."
The competition is open to teams from academic institutions, research organizations, university laboratories, and biotech and AI companies worldwide. Participants may use any models, algorithms or design methods. All submissions will go through standardized, double-blind experimental evaluation at MegaRobo's Megalaxy Lab.
Poreva Open is supported by Tencent Cloud, Twist Bioscience, Galatek, Vazyme, Guangzhou Pharmaceutical Digital Technology, Future Capital and Innoangel Fund, among others.
Five challenges and the metrics behind them
The inaugural competition consists of five challenges. The first three target EGFR, TNF-α and GLP-1R. Two more will be announced in the second round.
Experimental validation covers a range of metrics:
- binding affinity
- thermal stability
- cross-species reactivity
- pH selectivity
- nonspecific binding
- cellular function
Selected candidate molecules may advance to further testing in biological models, including organoids.
Results and experimental data, including data on unsuccessful designs, will be made publicly available after the competition, subject to intellectual property and compliance requirements. Publishing failed designs alongside successful ones is central to the benchmark idea, since it gives the field a record of what did not work in the lab.
Prizes and schedule
Poreva Open offers a total of US$250,000 in cash prizes (RMB 1.7 million). Megalaxy Lab will add US$150,000 (RMB 1 million) in wet-lab validation credits at no cost to participants. The top 10 participants may also have opportunities to secure up to US$1.5 million in potential investment.
- Oct. 19 to Nov. 18: registration and submissions for Challenges 1-3
- Nov. 9 to Dec. 8: submissions for Challenges 4-5
- By March 2027: experimental validation
- April 2027: results and experimental data published
- April 2027: team roadshow and awards ceremony
What it means for protein design teams
For teams developing AI protein design methods, the competition offers access to standardized wet-lab testing, with validation credits provided free of charge, and a common experimental standard for comparing approaches. Because the organizers plan to release the data after the event, the results could also serve researchers who did not take part, within the limits of intellectual property and compliance rules.
Registration is available through the Poreva Open website at porevaopen.megalaxylab.com. The organizers summarize the project with the line "Proof, not promises."
PR Newswire: 10,000 Protein Designs, One Wet Lab: Poreva Open Launches First Challenges
