This repository includes a vignette based on my presentation at BIRS Mathematical Frameworks for Integrative Analysis of Emerging Biological Data Types, June 2020.
The vignette covers the following topics:
- Using LIGER for the analysis of single cell transcriptomic data
- Visualizing data integration and clustering
- Utilizing metrics of clustering accuracy and integration success to compare analyses
Overview: In this analysis, the additional structure found by combining spatial and single cell transcriptomic datasets with an integrative nonnegative matrix factorization-based method, Liger, is demonstrated. From the separate and integrative analyses, plots of identified and known clusters are generated, metrics of integration performance are compared in context, and it is shown that there is some loss of information as a result of the integration.