Health tech giant Nvidia is collaborating with AstraZeneca and the University of Florida’s academic health center, UF Health, on new artificial intelligence research projects to accelerate drug discovery and improve patient care.
The organizations will use a new approach to training AI, called transformer neural networks, to allow researchers to leverage massive data sets, according to Kimberly Powell, vice president of Nvidia Healthcare, during a briefing with reporters. Transformer-based neural network architectures have become available only in the last several years.
Nvidia is collaborating with biopharmaceutical company AstraZeneca to build a model, trained using Nvidia DGX SuperPOD, an AI supercomputer, that will give researchers the ability to evaluate billions of molecules for potential drug candidates faster.
The MegaMolBART drug discovery model being developed by Nvidia and AstraZeneca is slated for use in reaction prediction, molecular optimization and de novo molecular generation. It’s based on AstraZeneca’s MolBART transformer model and is being trained on the ZINC chemical compound database, the companies said.
The large database allows researchers to pre-train a model that understands chemical structure and can be used for a number of downstream tasks, including predicting how chemicals will react with each other and generating new molecular structures.
It will be among the very first projects to run on Cambridge-1, which is soon to go online as the U.K.’s largest supercomputer.
"Just as AI language models can learn the relationships between words in a sentence, our aim is that neural networks trained on molecular structure data will be able to learn the relationships between atoms in real-world molecules,” said Ola Engkvist, Ph.D., head of molecular AI, discovery sciences and R&D at AstraZeneca. “Once developed, this NLP model will be open source, giving the scientific community a powerful tool for faster drug discovery.”
Chipmaker Nvidia also plans to collaborate with the University of Florida to develop an AI model called GatorTron that's trained using records from more than 50 million interactions with 2 million patients. That model will help identify patients for lifesaving clinical trials, predict and alert health teams about life-threatening conditions, and provide clinical decision support to doctors, Nvidia executives said.
"GatorTron leveraged over a decade of electronic medical records to develop a state-of-the-art model,” said Joseph Glover, provost at the University of Florida, in a statement. “A tool of this scale will enable healthcare researchers to unlock insights and reveal previously inaccessible trends from clinical notes.”
The clinical language model being developed with the University of Florida will help extract information out of massive untapped data sources like electronic health records, Powell said.
"We have built solutions addressing some of the computational challenges and opportunities in front of us. We are bringing the entirety of state-of-the-art computing to the healthcare industry," she said.
“The GatorTron project is an exceptional example of the discoveries that happen when experts in academia and industry collaborate using leading-edge artificial intelligence and world-class computing resources," said David Nelson, M.D., senior vice president for health affairs at the University of Florida and president of UF Health, in a statement. "Our partnership with NVIDIA is crucial to UF emerging as a destination for artificial intelligence expertise and development."
Beyond clinical medicine, the model also accelerates drug discovery by making it easier to rapidly create patient cohorts for clinical trials and for studying the effect of a certain drug, treatment or vaccine, the organizations said.
Famous for its consumer graphics cards used in the gaming market, Nvidia has been making moves into healthcare and life sciences. Its GPU chips that many use to play video games are also well suited in structure to power the massively parallel operations within supercomputers. Last year, the company announced plans to build an AI center of excellence in Cambridge, open to collaborations with U.K. researchers and startups, Fierce Biotech reported.
Also Monday, the company unveiled a central processing unit for giant-scale AI and high-performance computing applications that steps up its competition with Intel. The moves, outlined in the GTC keynote speech by Nvidia CEO and co-founder Jensen Huang, also solidify the company’s ambitions to be a leader in enabling AI and further transcend its legacy identity as a provider of graphic processing units, Fierce Electronics reported.