Pervasive AI

Artificial Intelligence and Machine Learning are transforming the life sciences in Indiana and beyond

Source: Indianapolis Business Journal's BioFutures Magazine

The concept of Artificial Intelligence has been around for decades, but 2023 was its coming out party.

Following Open AI's introduction of Chat GPT at the end of 2022, AI was on the tip of everyone's tongue. An analysis by NBC News found that mentions of AI during earnings calls skyrocketed in the first half of 2023. AI was mentioned on those calls as often as "interest rates" and "federal reserve," as companies as diverse as Salesforce and Walmart began hyping AI's transformative powers.

While companies of all kinds are talking about AI, its use isn't widespread. A November 2023 survey of more than 200,000 businesses conducted by the US Census Bureau found that only 4.4% of businesses were currently using AI.

But the adoption of AI is more prevalent in the life sciences, where the stakes couldn't be higher. AI and Machine Learning promise big advancements in drug discovery, clinical trials, precision medicine and other applications. As long ago as 2020, surveys showed that 90% of large pharmaceutical companies and 60% of emerging biopharma companies were deploying AI and Machine Learning.

Solving problems in research

AI's deployment in research is helping scientists overcome obstacles and unlock information in new ways.

At the Indiana Biosciences Research Institute, which facilitates research synergies between academia and industry, Director of Bioinformatics Travis Johnson is working with researchers at IBRI's Lilly Diabetes Center of Excellence to better understand the high-risk cell types involved in Type 2 Diabetes and guide development of novel targeted therapies.

The research involves "transfer learning," a type of AI that learns from one set of data and then applies it to another set where data is lacking. "In our case, we don't have diabetes risk information for individual cells, only for whole patients, so we infer this information at the single-cell level using transfer learning," Johnson said.

IBRI is also using AI tools to analyze nearly a million individual cells from the brains of obese and diabetic mice to study the metabolic effects on their central nervous system. "This task is only made possible via AI approaches that allow us to distill billions of gene sequences across nearly a million cells into usable visualizations and statistics tat give us actionable insights about the data," Johnson said. "We are already validating numerous genes identified from this study in the lab that may result in new drugs used to diagnose and treat obesity and diabetes."

To read the full article, go to the Indianapolis Business Journal BioFutures.