I have microarray gene expression data and their labels for a supervised tasks and some sample info, how can I estimate the batch effect?
what kind of gene expression data?
RNA-seq or microarray?
batch effect usually arises during analysis of cross-platform datasets
Microarray. I have updated the question
so you have microarray datasets from different platforms.
you can normalize their quantile wrt reference sample , and batch effect would be removed . Removed batch effect is estimation of batch effect.
If you are looking for a tool then COMBAT (http://www.bu.edu/jlab/wp-assets/ComBat/Abstract.html) is the answer.