Hello, I am investigating the output of chromVar, and I thought the z-scored deviation score normalizes the deviation score for each motif, and therefore the mean and variance of z-score matrix should be 0 and 1.
However, I tested both rowVars and colVars for the matrix, but the variance is not constantly 1. Could you explain why the variance of z-score is not 1?
This is the code I used:
computing deviations
dev <- chromVAR::computeDeviations(
object = chromvar.obj,
annotations = motif_ix,
background_peaks = bg
)
dev_score1 <- SummarizedExperiment::assays(dev)[[1]]
dev_score2 <- SummarizedExperiment::assays(dev)[[2]]
dev_score_z <- dev_score2
print(rowVars(dev_score_z))
print(colVars(dev_score_z))
And this is an example of row variance:
example of column variance:
Hello, I am investigating the output of chromVar, and I thought the z-scored deviation score normalizes the deviation score for each motif, and therefore the mean and variance of z-score matrix should be 0 and 1.
However, I tested both rowVars and colVars for the matrix, but the variance is not constantly 1. Could you explain why the variance of z-score is not 1?
This is the code I used:
computing deviations
dev <- chromVAR::computeDeviations(
object = chromvar.obj,
annotations = motif_ix,
background_peaks = bg
)
dev_score1 <- SummarizedExperiment::assays(dev)[[1]]
dev_score2 <- SummarizedExperiment::assays(dev)[[2]]
dev_score_z <- dev_score2
print(rowVars(dev_score_z))
print(colVars(dev_score_z))
And this is an example of row variance:
example of column variance: