Vanishing Gradients Episode 18: Research Data Science in Biotech
20 snips
May 24, 2023 Eric Ma, a leader in the research team at Moderna Therapeutics, discusses the tools and techniques used for drug discovery, the importance of machine learning and Bayesian inference, and the cultural questions surrounding hiring and management in research data science in biotech. They also explore the tech stack used in their work, the skills and hiring considerations in biotech, the importance of data testing and standardizing Excel spreadsheets, and the current state and challenges of Bayesian inference.
AI Snips
Chapters
Transcript
Episode notes
Feature Attribution Falls Short In Labs
- Explainable-AI feature attributions rarely satisfy wet lab scientists' needs.
- Biologists prefer mechanistic models or domain-specific visualizations like sequence logos tied to actionable levers.
Use ML When Experiments Are Costly
- Use ML-guided design when experiments are expensive because models reduce the number of required tests.
- Compare cost trade-offs: cheap random mutagenesis versus costly ordered variants before choosing ML.
Standardize On Differentiable Models
- Standardize on differentiable, simple, reusable neural architectures for sequence-to-function models.
- Optimize sequences by making models jointly differentiable so gradients guide efficient design.