(74a) Artificial Intelligence for Pharma 4.0: Challenges and Opportunities

In the pharmaceutical industry, the urgency to increase productivity, shorten development cycle time, reduce costs, and get it right the first time has never been more serious than it is now. The old empirical paradigm of guess-and-test, learning by doing, does not seem to be working as well as it used to. As the industry embraces its version of Industry 4.0 - Pharma 4.0 – to address these challenges, exciting recent advances in artificial intelligence (AI) offer important opportunities.

Decision-making in pharmaceutical product development and manufacturing involves the integration of process modelling tools, effective use of laboratory-generated information, use of knowledge from the scientific literature, as well as development of technical specifications and a knowledge base to satisfy regulatory requirements. The amount and complexity of information of different types, ranging from raw experimental data to lab reports to sophisticated mathematical models, that need to be stored, accessed, validated, manipulated, managed, and used for decision-making is staggering. Pharmaceutical discovery and manufacturing has long been a “Big Data” discipline. In fact, a typical New Drug Application (NDA) contains more than 100,000 pages of data, information, and knowledge.

To address the challenges, we need a systematic, integrated, informatics framework, pharmaceutical informatics, based on formal and explicit models of information. In addition, we also need tools that would support rapid extraction of mechanistic, first principles, knowledge from raw data gathered from PAT-like techniques. The information models need to be accessed easily by humans and software tools, and should provide a common understanding for information sharing. A critical challenge is the scarcity of semantically rich, properly populated, ontologies for the pharma domain.

Recent progress in machine learning, natural language processing, ontology engineering, etc., offers potential solutions to these challenges. In this talk, I will discuss a few case studies in structure-property prediction, hybrid-AI models, and a “Watson-like” system for pharmaceutical manufacturing.