OR WAIT null SECS
Their use in COVID-19 fight paves way for future leaps.
Artificial intelligence and computing companies around the world are currently applying their powerful technology to help scientists better understand and, hopefully, conquer COVID-19. Among them, IBM Research announced in April that it was making multiple resources available free to researchers, including a cloud-based AI
research resource that has been trained on a corpus of more than 45,000 scientific papers contained in the COVID-19 Open Research Dataset (CORD-19) and the IBM Functional Genomics Platform, a cloud-based repository and research tool built to discover the molecular features in viral and bacterial genomes. IBM Research has also played a leading role (in collaboration with the White House Office of Science and Technology Policy, the US Department of Energy, and others) in launching the US COVID-19 High Performance Computing Consortium, which is making available to scientists more than 38 US-based supercomputers with “more than 400 petaflops of computing power.”
“High-performance computing systems allow researchers to run very large numbers of calculations in epidemiology, bioinformatics, and molecular modeling, experiments that would take years to complete if worked by hand, or months if handled on slower, traditional computing platforms,” wrote Dario Gil, director of IBM Research, on the IBM Research blog (April 3). Dr. Teodoro Laino, manager of accelerated discovery, IBM Research Europe, explained IBM’s work in this area further. “Accelerating drug discovery does not come from the fact that you can extract data-you need to complement the data with physical and mathematical modeling, using high-performance computing, quantum, and AI,” he told Pharm Exec. “Once the data extracted from documents has been augmented and completed, these data can be used to create generative models, which can generate hypotheses. Following that is validation and testing.” IBM Research is now driving chemical labs completely on a system with no human intervention. Says Laino, “We’re relieving people from tedious programming tasks by having AI self-program the automation hardware for testing. It is a self-sustained, completely closed loop of accelerated drug discovery.”
If this drug discovery model currently seems not just beyond the capabilities of the individual, but also beyond that of many pharma companies, Laino reminds us that all the individual components mentioned “already exist in pharma R&D.” It is IBM’s integration of these technologies that serves to accelerate their power. Laino says that the quantum computing component will have an impact within this decade. “We will use quantum computing as the systems increase in their capabilities,” he explains, “for example, in their application to the calculation of properties of possible targets.” He adds, “We are also exploring the possibility of using quantum for some algorithms for designing retrosynthesis. This is a very complex problem; you have to explore a very complex graph of chemical reactions, and that’s where quantum technologies can actually make an incredible impact compared with classical computing.”
While, for some, quantum computing may conjure up thoughts of science fiction, Laino emphasizes that “the quantum computer is actually closer than people think.” He says, “You can easily have access to a quantum computer in 2020; in fact, it’s been possible to access a quantum computer since May 2016. The only constraint is that they work slightly differently to a traditional computer. Mapping the quantum architecture to the problem that you have to solve is part of the equation.” IBM’s largest quantum computing installation is a 53-qubit system on which “you can run molecules of a decent size,” says Laino, “but this isn’t yet comparable with what you can do on a high-performance computing installation.”
The situation is changing quickly, however. Laino believes it’s just a matter of years before “we will really see the gain from using quantum computers.” There are currently 18 quantum computers in existence; IBM is installing another one in Germany at the end of this year, followed by one in Japan. As part of the IBM Q Network, 110 companies can currently access these systems.
As well as being hopeful that the release of its resources will help to speed up the global effort against COVID-19, IBM’s Dario Gil wrote in April that the move is another step toward an age “where multi-disciplinary scientists and clinicians work together to rapidly and effectively create next-generation therapeutics, aided by novel AI-powered technologies.”
Julian Upton is Pharm Exec’s European and Online Editor. He can be reached at firstname.lastname@example.org