Balancing Automation and Experimentation in Biotech

In this episode of Data in Biotech, Ross Katz reflects on what he’s learned from one year of hosting the podcast. Diving deep into the intersection of data science and biotechnology, this episode covers topics like:

• The need for predictive models in biotech that are grounded in real-world experimentation.
• The challenges of biases in model evaluation and designing experiments that maximize collecting information for iterative improvements.
• The balance between leveraging computational methods and validating insights through experimental data. 

As we look to 2025, Ross shares his vision of the emerging democratization of the biotech data ecosystem by making domain knowledge, datasets and tools more accessible. He discusses the possibility of a future where decentralized collaboration, akin to open-source software projects, can tackle specific diseases through computational pipelines and cloud labs, enabling experiments without the need for costly infrastructure. Or where emerging trends like foundation models and ensemble modeling in drug discovery, cell and gene therapy, and the role of new data from advanced imaging and assay technologies can be unlocked to create novel insights.

Finally, he invites regular listeners to contribute ideas, guest suggestions and resources as we build community and embrace more curiosity and openness. 

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