Sunny Hung
Dan Price
Joe Boyer
Deepak Bandyopadhyay
Isha Bansal
Kate Adams
Kenneth Lind
Ninad Prabhu
Jaydeep Bardhan
Randy Smith
Jeffrey K Kerns
Meredyth Wolsky
Pharmaceutical companies are generating increasingly voluminous and complex biological and chemical data, which places heightened strain on proficiency with data at the project and portfolio levels alike. This strain is intensified in mature organizations that are heavily populated (particularly at leadership levels) with experienced scientists whose formal educations often pre-date contemporary advancements in data science & chemoinformatics. This gap between the workforce and modern data science leads to suboptimal utilization of data in decision-making.
Five years ago, several scientists at GSK, self-dubbed the Data Science Matrix Group (DSMG), sought to address this gap by hand-crafting a domain-relevant, hands-on educational curriculum to teach deliberately pragmatic data science skills to scientists with the intention of immediate translation to participants’ day jobs. The program runs one half-day per week for six months and emphasizes core competencies in Databases & SQL, Spotfire, Excel, Statistics, and Machine Learning fundamentals such as clustering. In addition, parallel curricula were offered to explore disciple-specific concepts, data sources and software. The Medicinal Chemistry track, for example, offers coursework in Chemoinformatics (SMILES, SMARTS, chemical similarity, chemical clustering, and diversity selection) as well as training on platform software such as LiveDesign.
This presentation will relate our experiences in producing and running the DSMG program, describe our means of assessing its success/merit, and convey some of our key learnings in evolving the content over 5 years. Over 100 chemists at GSK have participated in the DSMG program to date, including the entirety of US Discovery Medicinal Chemistry team/program leadership. The DSMG program has demonstrably improved data proficiency within the GSK Discovery organization and has provided a solid foundation for increased adoption of more advanced analytics and predictive modeling.
Ask a Question
Get involved to find out more about this Presentation.
All Comments
Log In to participate in the discussion
Discover more research and events on morressier.com