Starting Analysis from TPM matrix


I’d like to reuse this dataset for test some hypothesis.

The technology used is SMART-Seq2 and the matrix is already trasformed in TPM.

I have two question:

  1. Do you have some suggesting how to import in scanpy and avoid a further normalization step?
  2. My idea is to check the expression of gene in this human patients that is generally low expressed, MAGIC could be a good strategy?


  1. You should be fine to just read the csv and put the data into an anndata object. No normalization should be performed unless you apply it.
  2. Not sure what you mean by this. “Lowly expressed” sounds like you’d need a point of reference. If you’re just wanting to look at expression levels, why not just plot the expression levels for the gene?