When we use count data in RNA-seq analysis, we usually use normalization and sometimes vst, rlog transformation (DESeq2)or log2(CMP+4) transformation (edgeR) to perform K-means clustering. Can we use scaling after normalization and transformation? Or Should we use scaling only when we just use normalization without transformation?

  • 1
    $\begingroup$ Scaling in terms of Z-transformation? If so then yes, you can scale your log2 normalized counts to the Z score. This has the advantage that you unify the scale so all genes are represented by their deviation from the mean, regardless of the expression level. $\endgroup$ – ATpoint May 30 '20 at 8:53
  • $\begingroup$ I mean autoscaling and paretoscaling. $\endgroup$ – user224050 May 30 '20 at 14:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.