I have gene expression data from different conditions from different studies. Instead of using the actual TPM values for Pearson Correlation coefficient...