SandboxAQ Launches AQPotency, Bringing Ultrafast Virtual Screening to Drug Discovery Without a Solved Protein Structure

PALO ALTO, Calif., Aug. 19, 2026 /PRNewswire/ -- SandboxAQ today announced the general availability of AQPotency, its Large Quantitative Model (LQM) for predicting how well a potential drug will work, now available on Claude via Model Context Protocol (MCP). For any disease target, AQPotency scores how strongly candidate molecules are likely to act on it, and it does this without the expensive, time-consuming lab work that older methods depend on. Drug discovery teams can now rank enormous libraries of molecules by computer in seconds, then spend their limited lab budgets only on the candidates most likely to succeed.

Early drug discovery is a high-stakes guessing game. Teams have to decide which molecules to test in the lab, and a wrong call can cost months of work and significant expense. The established computer methods for narrowing the field are slow and costly to run, and they only work when scientists already have a detailed structural map of the disease target. Many of the most valuable targets have no such map, so those methods never reach them, and promising programs stall before they start.

AQPotency clears that roadblock. It runs on ordinary computing hardware, ranks pairs of molecules and targets in seconds, and costs as little as $1 per 1,000 comparisons. For every prediction, it also reports how confident it is and whether the target falls within the range where the model performs reliably. Older tools give scientists a single score with no sense of its reliability. AQPotency tells them not only what it predicts, but when they can trust it, which makes the results something teams can act on.

AQPotency also works in the other direction. Starting with a single promising molecule, it scans a broad panel of proteins across the body and returns a ranked list of the ones it is most likely to act on. When a molecule shows a useful effect but no one yet knows why, this gives research teams a focused set of leads to test.

"SandboxAQ's models have been very impactful for our work as we develop new treatments for Parkinson's," said Professor Dario R. Alessi, OBE, FMedSci, FRS, Director of the MRC Protein Phosphorylation Unit at the University of Dundee. "These models enable us to explore a much larger biochemical space in a short timeframe and improve both activity and selectivity. SandboxAQ's unique datasets and models stand out in the industry for their impact."

"This collaboration with SandboxAQ highlights the power of combining advanced AI-enabled discovery with rigorous experimental validation to unlock novel opportunities against historically difficult membrane targets. By identifying selective SV2C binders from a broad commercial library, the work establishes a compelling foundation for the development of first-in-class small-molecule tools and future therapeutics aimed at Parkinson's disease and other disorders of dopaminergic signaling," said Dr. Gary W. Miller the Adrienne Block Professor of Environmental Health Sciences and the Vice Dean for Research Strategy and Innovation at the Columbia University Mailman School of Public Health.

Andrea Bortolato, Vice President of Drug Discovery, at SandboxAQ, said: "AQPotency has given us and our customers a faster, scalable and reliable way to prioritize compounds in the workflows we already run, without needing a 3D crystal structure of the target. This opens up programs that structure-based methods simply couldn't reach. The confidence intervals make the output actionable for biopharma companies, and the model has already been successfully used in eight customer programs with experimentally validated impact."

"What's compelling about AQPotency is that it makes high-value discovery decisions faster and more practical," said Robin Roehm, CEO and Co-Founder at Apheris, which offers federated data and AI networks for life science companies. "Researchers can prioritize the most promising compounds with greater confidence, focus experimental resources where they matter most, and expand discovery efforts to targets that have traditionally been harder to pursue."

AQPotency is generally available today through Claude via MCP and SandboxAQ's website, with availability on Google Cloud's Marketplace to follow. Its general availability is paired with the GA on Claude and Claude Science of a second LQM, AQCat Adsorption Spin, for catalyst discovery.

About SandboxAQ

SandboxAQ is a B2B company delivering solutions at the intersection of AI and quantum techniques. The company's Large Quantitative Models deliver critical advances in life sciences, financial services, navigation, and other sectors. SandboxAQ is an independent, growth backed company funded by leading investors and strategic partners including funds and accounts advised by T. Rowe Price Associates, Inc., Google, Alger, IQT, US Innovative Technology Fund, S32, Paladin Capital, BNP Paribas, Eric Schmidt, Breyer Capital, Ray Dalio, Marc Benioff, Thomas Tull, and others. For more information, visit www.sandboxaq.com.

SOURCE SandboxAQ