This is a test version of Biostars. For the public version, visit https://www.biostars.org.
log2 transformation in the GEO2R script for microarray analysis

Hello,

In the GEO2R R scripts to analyse microarray data (https://www.ncbi.nlm.nih.gov/geo/info/geo2r.html), there is an auto-detect feature to determine if the values are or are not in log space:

ex <- exprs(gset)
qx <- as.numeric(quantile(ex, c(0., 0.25, 0.5, 0.75, 0.99, 1.0), na.rm=T))
LogC <- (qx[5] > 100) || (qx[6]-qx[1] > 50 && qx[2] > 0) || (qx[2] > 0 && qx[2] < 1 && qx[4] > 1 && qx[4] < 2)
if (LogC) {
    ex[which(ex <= 0)] <- NaN
    exprs(gset) <- log2(ex)
}

My question is, how these specific quantile conditions (i.e. qx[5]>100) can determine if the values are or not in log space?

Thanks a lot

r
• 4,331 views • link
updated by Ram Ram • written by faustinomarin10 • faustinomarin10

1 answer

if qx[5] > 100 is true than at least 1% of the data is bigger than 100. A value of 100 or bigger can not be the result of log transformation because the reverse 2-log (2^100) is so big that you cannot even calculate it!
• 0 views • link
written by Irsan Irsan

I agree with your reasoning. But, why specifically 100? I wonder if it is arbitrarily chosen.

Also, concerning the values for the other quantile conditions, I think other different explanations would be needed.

• 0 views • link
updated by Ram Ram • written by faustinomarin10 • faustinomarin10

It's a nice round number, 99 or 101 would work just as well. That's also why 50 is used in testing the range.

• 0 views • link
written by Devon Ryan Devon Ryan

Log in to answer this question.

More posts like this