Causal Inference in Single-cell Genomics - Machine Learning in Computational Biology

Explore causal inference in single-cell genomics through this research seminar presented by Yongjin Park from the University of British Columbia. Delve into the unique aspects of single-cell RNA-seq data and discover how statisticians can leverage its special structure. Learn about a novel algorithm for ascertaining the effect of disease status on cell-type-specific gene expression profiles, and gain insights into various causal effect inference strategies. Examine scalable approaches for cell type assignment and the integration of single-cell data with existing tissue-level bulk data. Uncover how this integrative analysis provides high-resolution, cell-type-level views of complex disease mechanisms in genome-wide association studies. The seminar covers topics such as differential expression analysis, cellular context, biological covariates, single-cell mixture models, and self-annotation techniques.

Introduction
Outline
Causal Inference
Notations
Observations
Assumptions
Logistic regression
Randomized control trial
Inverse probability weighting
Causal assumptions
Summary
Differential Expression Analysis
Cellular Context
Biological Covariates
Singlecell Data
Pipeline
Pipeline of Singlecell Analysis
Singlecell Mixture Model
T cell study
Self annotation