Should MAGeCKFlute CellCycle Normalization make beta scores more extreme?
Entering edit mode
Solarion • 0
Last seen 7 months ago
University Hospital Jena, Germany

I used MAGeCKFlute downstream of MAGeCK MLE to normalize the beta scores by the cell cycle, as recommended by the package documentation. However, the beta score distributions from both my conditions do not shift but become more extreme. In the Flute paper the given examples were beta-score distributions that are centered around -1 and after normalization became centered around 0, but with the same distribution width. For me, my distributions are already centered around 0, and when I normalize the only thing that changes is that the distribution width increases (from a minimum of about -2 to a minimum of about -5). This happens in both conditions equally (the scatterplot mostly changes in scale, not so much in how the points are located towards each other).

Question no 1: is that behaviour expected and normal? Question no 2: if so, do I still apply the same static absolute threshold to call dependencies? (e.g. -0.5) This would lead to a very high number of dependency-genes where there were only a handful before.

My concern is to call false positive targets from the CRISPR screen, due to artificially extreme beta scores.

My code:

edata1 = ReadBeta(file1)

edata2 = NormalizeBeta(edata1, samples=c("Ctrl", "Treat"), method="cell_cycle")

MAGeCKFlute • 57 views

Login before adding your answer.

Traffic: 444 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6