AI-powered digital innovations in pharmaceuticals research & development: Current landscape and case examples

Artificial intelligence (AI)-powered digital innovations are increasingly reshaping pharmaceuticals and research and development. While AI/ML techniques are now widely accessible, some of the work on Large Language Models (LLMs) and other aspects of digital innovations like spatial omics are just beginning to take off; thus, recent learnings are particularly useful to help statisticians and data scientists better understand the field. This paper examines current trends in AI-powered digital innovations within the pharmaceuticals industries, focusing on the challenges and opportunities associated with LLMs, other Generative AI (GenAI) approaches, and advanced analytical methods such as ML models for spatial omics. We review recent applications of these technologies by leading pharmaceutical companies, highlighting both the potential benefits and the inherent challenges and opportunities associated with their adoption. We will briefly mention recent regulatory guidance/reflection paper on the development of AI-driven tools in drugs and devices. To illustrate clinical and operational utilities, we present case examples from industry, detailing the adoption of AI in high-plex tissue image analysis through deep learning techniques for automated single-cell segmentation and spatial pattern discovery across omics datasets, and demonstrating how Retrieval Augmented Generation (RAG) approach of using LLM may help to standardize clinical trial monitoring. By highlighting the above, this paper aims to provide statisticians and data scientists in the pharmaceutical industry an insightful overview for integration of AI in their day-to-day R&D and generate ideas on future applications.