I have a question regarding the MEDIPS.meth output. I have 2 data sets (no replicates) and I would like to get the differential methylation coverage between them. I undestand that the log2FC are calculated: log2 (Mset1-Mset2) but when I check the output I see that in some cases I got log2FC<0 when the counts are higher in Mset1 than in Mset2. So I wonder how is it calculated? using normalised counts?

Code should be placed in three backticks as shown below

mr.edgeR <-  MEDIPS.meth(MSet1 = bam_clon,
                         MSet2 = bam_wt,
                         CSet = CS,  #COUPLING SET
                         p.adj = "bonferroni", 
                         diff.method = "edgeR", 
                         MeDIP = T, 
                         minRowSum = 10,
                         CNV = F,
                         diffnorm = "tmm")


# and this is the output

mr.edgeR[105:106, c(1:10,15:20)]

     chr start  stop CF H827_12.bam.counts H827_wt.bam.counts H827_12.bam.rpkm H827_wt.bam.rpkm H827_12.bam.rms H827_wt.bam.rms
105 chr1 10401 10500  6                 14                  9             1.90             1.44            0.29            0.25
106 chr1 10501 10600  7                  6                  5             0.81             0.80            0.22            0.21
    MSets2.rpkm.mean MSets2.rms.mean edgeR.logFC edgeR.logCPM edgeR.p.value edgeR.adj.p.value
105             1.44            0.25  0.29504070    -4.564983     0.6778166                 1
106             0.80            0.21 -0.07439969    -5.406249     1.0000000                 1

#In lane 105 with Mset1 counts > Mset2 counts I get a log2FC > 0
#In lane 106 with Mset1 counts > Mset2 counts I get a log2FC < 0

I´m guessing there is some normalisation underneeth that I am not aware of and that would be the reason for that.. I would appreciate your help with that. Thanks!!