TCGA RNAseqV2 data matrix batch effect control and remove

Question

[fractal61]

I have a RNAseqV2 data matrix, rows are genes and columns are samples and cells values are raw counts. I also have normlized data. Dimensions of data 18323 x 600. How can i check batch effect from this data and remove? Is it necessary process for obtaining differential expressed genes.

Answer 1

[andrew.j.skelton73]

Use principle components analysis (PCA) to try and eyeball potential batch variables - You could also look at the SVA package too. Correcting for batch effects depends on the situation, but I'd advise adding the variable in as an additive term in the linear model design. See limma