Can or should we do batch-correction before parsing the log2cpm to CONICSmat?

[biobug16]

Hi @diazlab and @soerenmueller I want to do the batch-correction for my samples before predicting CNVs using CONICSmat. Do you recommend any method for batch-correction as my histogram for the Zscore and for the posterior probabilities, do not represent few of my samples properly. Will batch-correction help in this regard. Is there any step within CONICSmat which takes care of this issue and may be I am missing it? Any suggestion will be appropriated. Thanks