(423d) Towards a Comprehensive Ontology to Describe a Pharmaceutical Product | AIChE

(423d) Towards a Comprehensive Ontology to Describe a Pharmaceutical Product


Viswanath, S. - Presenter, Eli Lilly and Company
Vaidyaraman, S., Eli Lilly and Company
Guntz, S., Eli Lilly and Company
Dieringer, J., Eli Lilly and Company
Wang, H., Carnegie Mellon University
Gounaris, C., Carnegie Mellon University
The design of pharmaceutical chemistry manufacturing & controls (CMC) ontologies to fully describe a pharmaceutical product hold significant business value, given the multiple applications in efficient resource and activity planning, accelerating submissions, and building a core digital foundation for a multitude of analytics applications. The flexibility afforded by the inherent data structure of an ontology affords the adoption of an 'agile development' paradigm, wherein targeted data sub-sections are captured, while others are captured later without compromising the comprehensiveness of the solution. This talk will initially showcase a top view ontology based on product requirements, customers, decisions, activities and resources. A specific example of such an ontology instanced for multiple small and medium molecule assets will be shared. The middle layer of the ontology is based on risk-based development, and the integration of a comprehensive array of risk grids to the top ontology layer. The integration of these risks into the ontology allows for a dynamic digital connection of CMC risks directly to product requirements (that come from patients, payers, providers, manufacturing, clinic and the business), and enables powering decision making (activity selection, resource allocation) to optimally reduce CMC risks subject to constraints. Lastly, specific risk grid attributes within the risk grids are linked to instrument data (powered by instrument-specific ontologies such as those from Allotrope) to build the connection to the specific data elements. A case study showcasing the power of setting up product requirements, that lead to decisions to commission activities which consume resources, but also reduce the CMC risks by generating specific data, will be shared. Longer term plans to design and install these ontologies is a pre-competitive collaboration topic across pharma and will be discussed