Batch characterization
Batch effects and data integration are well known challenges in scRNAseq. Still it’s not well known how differences between batches/datasets are reflected within the datastes features.
Open questions are:
- Are batch effects/differences celltype specific?
- How can we determine the size/strength of the differences?
- Can we separate batch effects into groups of “batch types”?
- Can we identify a subset of “batch genes” that are solely/specifically affected?
Here we analyzed datasets with different sources of known batch effects and characterized their batch effects according to these questions.
Results of the analysis are shown here:
