Background

• Overview of the scvi-tools codebase
• Counterfactual prediction
  • Preliminaries
  • In-distribution vs. out-of-distribution
  • Applications
• Differential Expression
  • Problem statement
  • Motivation
  • Notations and model assumptions
  • Approximating population-specific normalized expression levels
  • Detecting biologically relevant features
  • Providing easy-to-interpret predictions
• Transfer learning
  • Reference mapping with scArches
• Variational Inference
  • Approximating the posterior
  • End-to-end learning
  • Amortizing inference