Generative AI for Modeling Single-Cell Responses

Explore a comprehensive lecture on generative AI applications in single-cell response modeling presented by Fabian Theis at the EWSC-MIT EECS Joint Colloquium Series. Delve into topics such as engineering considerations, generative modeling techniques, data integration strategies, and representation learning. Examine the creation of an integrated lung cell atlas and learn about distribution learning, population-level integration, and phenotype mapping. Investigate experimental design, results analysis, and the incorporation of priors in modeling. Discover the potential of differential programming and multiomics approaches, including practical examples and human prediction models. Gain insights into cell embedding techniques and engage with a Q&A session covering various aspects of the presented research.

Introduction
Engineering matters
Generative modeling
Data integration
Representation learning
Integrated lung cell atlas
Learning distribution
Population level integration
Scale
Map phenotype
Multigrade
Experimental design
Experimental results
Adding priors
Differential programming
Multiomics
Multiomics example
Prediction human
Embedding cells
Munich
Questions
Encoder
Question