Micromanufacturing is on the cusp of providing more affordable and more accessible therapeutics, especially for orphan indications, which require only small, finite quantities.
Sqale, incubated at the BioInnovation Institute in Denmark, is pioneering a method, dubbed BioBox™, that brings a faster and more affordable approach to modular, decentralized biomanufacturing by using what co-founder Seyed Mansouri, PhD, calls “quantum generative algorithms.”
Possible applications include making small indication therapeutics and vaccines, such as HIV and Zika vaccines, as well as gene therapies and anti-venoms. In such cases, “If suitable manufacturing isn’t in place, the cost will be so high we can’t make it,” Mansouri says.
The goal is to create a “self-driving,” data-driven manufacturing facility that can be controlled remotely, without the need for PhD-level scientists onsite. This way, lower volume treatments such as gene therapy can be received by those who need them rather than only by privileged populations.
“In today’s world, if suitable manufacturing isn’t in place, the costs for these biologics are so high they can’t be produced for millions of patients who are struggling with these conditions,” Mansouri tells GEN. The problem is that the necessary quantities are too small for current manufacturing practices to be cost-effective, or that remote locations lack the infrastructure for safe delivery.
Therefore, Sqale’s founders are rethinking therapeutic manufacturing from the ground up. As he points out, “The pharmaceutical industry still uses quite a bit of recipe-oriented technology, which requires massive amounts of resources and can be quite wasteful. Billions of dollars have gone to drug innovation, but relatively little has gone to rethinking and innovating manufacturing. Yet manufacturing is mission-critical.”
Industrial generative AI
“About 18 months ago, Zapata AI, which spun out of Harvard University’s quantum computing and chemistry labs in 2017, approached us. It was looking to apply industrial generative artificial intelligence (AI) algorithms enhanced with quantum methods to pharmaceutical manufacturing,” Mansouri recalls.
Basically, it applies “quantum-based algorithms to generate models in industrial use cases to extrapolate from known data to generate high-accuracy datasets by learning situation-specific rules in novel regimes,” he explains. The resulting virtual sensors are being used by Insilico Medicine and St. Jude Children’s Research Hospital to speed drug discovery and quickly identify oncology drug candidates with the highest potential.
Outside of medicine, INDYCAR® auto racing giant Andretti Global uses this approach to virtually measure the slip angle of race cars—with 99.5% accuracy—during races to manage tire degradation, predict yellow flags, and support pit stop strategy.
Sqale is using quantum generative AI to future-proof manufacturing. Within a few decades, Yudong Cao, Zapata co-founder and CTO predicts quantum hardware devices will be instrumental in drug discovery.
When that happens, “Our manufacturing will be ready, unlike other pharmaceutical manufacturing technology. You can’t retrofit quantum AI-models, and digital-twin assisted manufacturing into existing technologies,” Mansouri says. “Instead, you need to prepare for the next computational and digital paradigms. You need to rethink manufacturing from the bottom up.”
