MSstatsTMT Global Median Normalization

19 views
Skip to first unread message

Zachary Hutchins

unread,
Apr 4, 2025, 9:00:35 PMApr 4
to MSstats
Hi, 

I just wanted to confirm how the global median normalization is equalizing the medians. When equalizing the channels to the global median, is a channel being multiplied or divided by a value or is it having a value added or subtracted out of it?

Thanks,
Zachary Hutchins

Devon Kohler

unread,
Apr 9, 2025, 9:23:35 AMApr 9
to MSstats
Hi Zachary,

The channels are equalized by subtracting the differences between medians so that the channels median values align. You can see the code below for more info!

Best,
Devon

.normalizePeptides = function(input) {
    input[, MedianLog2Int := median(log2Intensity, na.rm = TRUE),
          by = c("Run", "Channel")]
    median_baseline = median(
      unique(input[, list(Run, Channel, MedianLog2Int)])[, MedianLog2Int],
      na.rm = TRUE
    )
    input[, Diff := median_baseline - MedianLog2Int]
    input[, log2Intensity := log2Intensity + Diff]
    input[, Intensity := 2 ^ log2Intensity]
    input[, Diff := NULL]
Reply all
Reply to author
Forward
0 new messages