Question: How can I remove batch effects among biological replicates? Does
cellranger aggr correct for batch effects?
Answer: Currently, the
cellranger aggr pipeline only performs read depth normalization among the 10x libraries combined.
In general, normalization and correcting for batch effects in single-cell RNA-seq data are areas of active research, and at present, we don't have a specific recommendation.
Below are some examples of performing batch correction in Seurat:
- The following tutorial shows an example using the RegressOut function:
- Alternatively, below is a tutorial for “aligning” data sets or samples which are expected to be similar (as in the case of biological replicates):