LLMs and Generative AI for Healthcare

Chapter 5. LLMs in Pharmaceutical R&D, Public Health, and Beyond

Drawing on the groundwork laid by the previous chapters discussing patient-facing and clinical use cases, this chapter now expands the horizon. We will discuss how large language models (LLMs) can be deployed to accelerate the discovery of therapeutic drugs by exploring the scientific literature and finding promising molecules. By examining a range of use cases, we will explore how exactly LLMs and generative AI can help different healthcare stakeholders in fields such as pharmaceutical research and public health.

Pharma Research and Development

The very nature of human biology can make it difficult to decipher, and finding new drugs reflects that complication. Each claim may require a decade or more to verify, and each one costs at least a billion dollars—on average, 10–15 years and more than $1 billion to bring one new medicine¹ through a successful journey from laboratory to pharmacy shelf. Researchers often go through sleepless nights and years trying again and again to reach their goals when others doubt them, hoping to find ways to transform the course of life-limiting or fatal diseases, and to ultimately give more patients the healthy lives they deserve.

Equipped with multifaceted talents and well-developed tools, various human scientists become like molecular architects (Figure 5-1) building thousands of brand-new molecular objects. Drug development is also divided in several stages: preclinical, clinical, ...