add script with leukdata processing in inst/extdata
Closed this issue · 3 comments
Would be really nice to see how you processed the Golub data.
It doesn't change your results, but right now someone else is not going to get the same leukdata as you are without this script or Rmd.
I totally agree — I recently also received an email request for this — in a 😞 for reproducible research, this was cleaned before I joined so I don’t have that script, but it’s on our radar & I am working to track it down.
Yikes! I hope you find it. I know I followed some stuff from the Bioconductor tutorial on it, and the max value they recommended was 16000, which on log10 scale would be 4.2, but the max I see in leukdata is 3.8, so very odd.
And I know we are dealing w/ roughly the same set of genes, as Golub_Merge from golubEsets 📦 has 7129 genes, vs the 7128 in leukdata.
The script has been located! Please find it here. Essentially it was normalized based on the methods described in Efron's Large Scale Inference (2010), section 6.1, page 91.