How can AI advance healthcare? The question is top-of-mind for everyone in the industry. A tsunami of data is powering AI’s potential. In 2023, for example, GSK generated more data in a single quarter than in the company’s previous 300 years of history. Michael Montello, MBA, GSK’s Senior Vice President for R&D Digital and Tech Tech, shared his insights into the top six areas of AI’s impact on healthcare, building big data and how AI is changing the nature of work in drug discovery and manufacturing.
Accelerating drug discovery. AI combined with genetics and genomics can validate targets, which is the starting point of developing new medicines and vaccines. Validated targets are twice as likely to succeed in clinical development. The other side of the coin: with genetic and genomic data, AI can help identify patients who will respond best to new drugs, making clinical trials more efficient and improving the probability of success.
Digital twins. When interrogated with AI, digital representations of cells or organs composed of hundreds of billions of data points can help understand biology and the root causes of disease. More important than the number of biomedical data points is their connections and semantic meaning. AI machine learning engineers can now see connections that could not be revealed using other tools before AI. Digital twins can also work in manufacturing, for example, to digitize the process of making a molecule and tune that process to increase yield.
Predictive power. By using predictive data sets from companies, academia and government working together, AI has the potential to predict infectious diseases that are on the horizon and provide early warning systems for pandemics.
Sharing large data sets. Big pharma created all its own data in the past, and much is still generated internally. To harness AI, partnerships with other companies, including tech companies, can expand data access. For example, GSK can access oncology data through a partnership with Tempus.
Good data governance. Using AI for good in healthcare, and avoiding disruption, requires big investments in data management and ensuring data provenance, lineage, quality and security.
Staffing up. Leveraging AI for good in healthcare depends on people—a new breed of collaborative techno-biologists. At GSK, this includes talented people from disciplines like data engineering, as well as interdisciplinary teams with collaborative mindsets focused on patient data, genetics and genomics, chemistry and manufacturing control.
“We need to have technology, science and talent together solving these problems,” says Montello. “Historically, science was super important, and technology was the enabler. Now technology and science are working together in ways that are much closer and more aligned with the big problems that we’re solving. It’s a big shift in ways of working, being inclusive to the multidisciplinary approach of collaboration.”