[Image: 1910 Genetics]
Microsoft and 1910 Genetics have announced a partnership that aims to reverse the troubling trend.
Accelerating discovery with AI and quantum-inspired computing
[1910 Genetics]
By focusing on democratizing these powerful technologies, the Microsoft–1910 pact is not aiming at incremental R&D improvements. The goal of Azure Quantum Elements is to “compress the next 250 years of chemistry and materials science progress into the next 25,” noted Microsoft CEO Satya Nadella when debuting the platform in 2023.
Microsoft–1910 Genetics partnership aims for breakthroughs in weeks, not years
Jen Nwankwo, Ph.D., CEO of 1910 Genetics
The potential of this approach is already evident in material science. In collaboration with the Department of Energy’s Pacific Northwest National Laboratory (PNNL), Azure Quantum Elements played a key role in a significant breakthrough: the discovery of a promising new solid-state battery electrolyte using Azure’s AI models and HPC capabilities, researchers rapidly screened over 32.6 million material candidates, narrowing them down to 18 top prospects for experimental validation in just a week. This resulted in a new electrolyte using 70% less lithium than traditional options, discovered in less than nine months.
The promise of Azure Quantum Elements could be similar in drug discovery, said Jen Nwankwo, Ph.D., CEO of 1910 Genetics. “I think we’re going to get higher quality drug candidates,” she added, and with fewer iterations. She believes this partnership will help the industry “jumpstart” the transition from “a traditional human-intensive, rational drug design approach to one that is AI-enabled.”
Targeting a billion-dollar business in AI-driven drug discovery
This collaboration also has the potential to reduce costs by using AI to identify drugs unlikely to succeed early in the process while also finding a variety of Azure Quantum Elements tools to support researchers across various stages of drug discovery. “Together with 1910, Microsoft is charting a course for a billion-dollar business in AI-driven drug discovery and development,” said Zulfi Alam, Microsoft’s vice president of quantum.
Microsoft–1910 Genetics partnership prioritizes faster drug discovery and higher-quality candidates
Nathan Baker, Ph.D.
The partnership between 1910 and Microsoft Azure Quantum Elements will combine 1910’s computational and wet lab biological data, robotics-driven laboratory automation and multimodal AI models for drug discovery with Microsoft Azure Quantum Elements. Nwankwo expects the partnership to lead to “significant gains” in accelerating the drug discovery process. She believes “we’re going to get higher quality drug candidates because we’ll need fewer iterations to get to the final molecule that makes it into the clinic and, hopefully, onto the market.”
Microsoft’s Nathan Baker, Ph.D. echoes this potential, emphasizing the goal of making powerful tools accessible in chemistry: “Can we make it easy for people to use AI and HPC together in the same platform,” said Baker. Microsoft envisions the potential of a supercomputer for every chemist. “We’d like to make it very straightforward to spin up the environments you need to do world-class science using best in class HPC and AI in the cloud,” Baker said.
Azure Quantum Elements aims to accelerate scientific discovery by uniting AI, cloud-based high-performance computing (HPC), and eventually quantum computing as the technology matures. “We see this as an onramp to future quantum computing,” Baker said. This emphasis on combining AI and HPC, along with streamlining scientific workflows, aligns directly with 1910 Genetics’ mission in drug discovery.
While quantum computing and AI are developing on somewhat parallel tracks, they both can accelerate material science research in complementary ways. “AI speeds things up. Quantum is going to help us with the accuracy,” Baker said. Their integration into a cloud-based HPC platform supports scale.
How ‘copilots’ can remove roadblocks
Azure Quantum Elements also promises to transform drug discovery through tools called ‘copilots,’ a topic Microsoft’s global business leader for health and life sciences, also described in a recent AI Meets Life Sci podcast. These tools streamline complex research processes, allowing scientists to focus on finding breakthrough treatments rather than technical hurdles. Copilots facilitate scientific processes by simplifying the creation of workflows, thus removing barriers to using advanced computational tools and allowing scientists to concentrate on their research questions without getting bogged down in the technicalities of software and workflows.
“If you think about drug discovery processes, they never involve just one calculation. There’s multiple steps. You screen for properties. Then you look at post prediction for a drug on the protein, then you do another calculation to understand binding affinity. But it’s all of these many calculations,” Baker said. These copilots simplify the creation of complex workflows, removing technical barriers and allowing scientists to concentrate on their core research questions. Consider drug discovery: it involves multiple calculations and tools. Copilots can automate the process of connecting these tools and building the necessary workflows. This removes the technical complexities, freeing scientists to focus on the science itself and to foster innovation in their research strategy.
The competitive landscape in quantum and quantum-inspired computing for drug discovery is rapidly diversifying. Startups like Sandbox AQ and Qubit Pharmaceuticals, alongside tech giants like NVIDIA and Fujitsu, are driving innovation. Collaborations between pharmaceutical leaders like Sanofi and Roche with tech innovators like Aqemia and Quantinuum further demonstrate the industry's momentum.
Microsoft’s technology transfer vision
This growing interest underscores the transformative potential that Microsoft sees in its partnership with 1910. In 2023, 1910 Genetics was the sole biotech to get a private preview of Microsoft Quantum Elements along with material science and chemical engineering giants BASF, AkzoNobel, AspenTech, Johnson Matthey, and SCG Chemicals. The 1910 private preview led to a commercial partnership in which 1910 will “move its entire platform on top of Azure Quantum and go to market globally,” Nwankwo said.
Baker's vision for Azure Quantum Elements emphasizes democratizing access to powerful tools: "At Microsoft, a lot of what we are doing is technology transfer. We take artificial intelligence models that were at the research stage, test them, harden and deploy them so that others benefit from them.”
Three collaboration models in drug development
Interested drug developers can choose between co-discovery, co-engineering or platform-as-a-service approaches.
In the co-discovery option, a pharma company comes to 1910 and Microsoft and provides selected disease targets. “Then we'll do the drug design on the Azure Quantum Elements platform, manufacture drugs, test them, and then deliver those molecules," Nwankwo said.
The co-engineering model involves building AI models and generating large-scale datasets tailored to in-house drug discovery efforts. Nwankwo emphasizes the value of this option for generating “wet-lab ground-truth biological data.” She explained, “We've gotten a lot of pharma companies coming to us and saying, ‘Hey, we're interested in how you generate a computational data stream. Can you teach us how to do that?’ Under this co-engineering model, we'll be helping partners create these types of datasets and helping them build AI models that they can run locally in their organizations.”
"It's like ChatGPT for enterprises," Nwankwo said. "You have the software and modules. With the click of a button, it can do drug design for you. You can run modules around predicting toxicity or other kinds of metabolic liabilities of your molecules and so on.”
Nwankwo said that the Microsoft–1910 alliance could set a new standard for collaborative efforts between tech and pharma. "I know you've seen other kinds of collaborations that are pretty light on detail. What we're saying here is that this is not that.”
Filed Under: Biologics, Biotech, Data science, Drug Discovery and Development, Genomics/Proteomics, Industry 4.0, machine learning and AI
