Tutorial

• Tutorial 1: prepare the preprocessed scRNA-seq data
  • conduct gene alignment with the Panglao dataset
  • normalize the aligned scRNA-seq dataset
  • save the pre-processed datasets
• Tutorial 2: run CANAL with data streams from various batches
  • construct the CANAL model
  • stage 1
  • stage 2
  • stage 3
  • load the unlabeled test data and apply CANAL to predict cell types
  • evaluate the annotation performance
• Tutorial 3: run CANAL on cross-tissue experiments
  • construct the CANAL model
  • stage 1
  • stage 2
  • stage 3
  • stage 4
  • load the unlabeled 10X test data and apply CANAL to predict cell types
  • evaluate the annotation performance
• Tutorial 4: detect novel cells in the test data via CANAL
  • Construct and run the CANAL model for three stages
  • predict cell types of the test data
  • evaluate the performance of cell-type annotation and novel cell detection