How Bayesian Optimization is Helping to Accelerate Innovation at Merck Group

Feb 14, 2024

In this podcast, Wolfgang from Merck discusses BayBE, an open-source library for Bayesian optimization, its benefits, challenges, and the need for standardized data models. He shares insights on accelerating innovation, experimental design, and advice for scientists in the industry.

Ask episode

Chapters

Transcript

Episode notes

Introduction

00:00 • 2min

Data Science and Bayesianformatics at Merck Life Sciences

01:32 • 4min

Advancing Experimental Design with Bayesian Optimization

05:29 • 10min

Balancing Objectives in Experimental Design with Bayesian Optimization

15:22 • 2min

Enhancing Efficiency and Collaboration Through Open-Sourcing

17:46 • 4min

Optimizing Efficiency through Bayesian Optimization and Transfer Learning

21:55 • 15min

Exploring Interdisciplinary Learning and Problem-Solving in Biotech

36:42 • 3min

This week's guest is Wolfgang Halter, Head of Data Science and Bioinformatics at Merck Life Science, a leading global science and technology company.

Ross sat down with Wolfgang to discuss the work on the BayBE project, an open-source library built for Bayesian optimization. Throughout the episode, we go on to learn how BayBE is used for both experimental design and as a means to accelerate innovation. The pair also discusses the benefits and challenges of Bayesian optimization and the need for standardised data models. Finally, Wolfgang shares some advice for those scientists and engineers who are keen to get ahead in the industry.

You can access the GitHub repo mentioned in the episode by clicking here: github.com/emdgroup/BayBE

Data in Biotech is a fortnightly podcast exploring how companies leverage data innovation in life sciences.

Chapter Markers:

[1:32] Wolfgang gives us a whistle-stop tour of his career to date and explains the motivation behind pursuing a career in Data Science.

[2:35] Ross asks Wolfgang about Merck’s mission and the role the data science team is playing in helping the company achieve that mission.

[5:28] Wolfgang explains the work that is going into the BayBE project.

[13:23] Ross asks Wolfgang how Merck arranged their experimental campaigns in BayBE and how they garnered insights during the process.

[17:45] Wolfgang explains why the team developed BayBE as an open-source library.

[19:25] Wolfgang shares some more details on how the data science team at Merck is using BayBE today.

[20:42] Wolfgang shares some examples of the kinds of applications that the team is currently developing.

[21:54] Wolfgang provides us with information about the amount of time that is saved on average as a result of adopting this approach.

[34:38] Ross asks Wolfgang how his engineering background informs his perspective on the problems facing biotech and R&D.

[36:57] Wolfgang gives us his advice for young scientists and engineers who are looking to learn more about biotech.

[38:24] Wolfgang provides us with a list of resources for those who want to find out more about Merck and the BayBE project.

Download our latest white paper on “Using Machine Learning to Implement Mid-Manufacture Quality Control in the Biotech Sector.”

Visit this link: https://connect.corrdyn.com/biotech-ml