I have a little question about dmc analysis with methylKit. I have a control sample as:
Chromosome 894 - 34 181 0.15814 CHG 2.07E-55 2.15E-53
Chromosome 895 + 44 161 0.214634 CHG 1.32E-77 4.16E-75
Chromosome 1866 - 14 95 0.12844 CHG 2.27E-22 1.03E-20
Chromosome 1998 + 32 161 0.165803 CHG 5.89E-53 5.54E-51
Chromosome 2041 - 31 115 0.212329 CHG 5.33E-55 5.45E-53
Chromosome 2339 - 23 120 0.160839 CHG 3.51E-38 2.02E-36
Chromosome 2612 + 25 122 0.170068 CHG 5.10E-42 3.26E-40
And treatment sample as: Chromosome 894 - 17 181 0.0858586 CHG 4.58E-24 8.34E-22
Chromosome 895 + 10 160 0.0588235 CHG 4.37E-13 2.64E-11
Chromosome 2041 - 10 132 0.0704225 CHG 7.13E-14 4.55E-12
Chromosome 2339 - 11 164 0.0628571 CHG 1.42E-14 9.51E-13
Chromosome 2612 + 16 145 0.0993789 CHG 8.75E-24 1.54E-21
As you see, site 1866 and 1998 were only detected to be methylated in the control sample. When I use the unite command to merge the data, these two sites were not included in the merged data, since they only existed in the control sample. And then they will not appear in the final dmc result.
So, is there a way that I could make sites like this also be included?
I have tried to firstly mark both samples as control, merge them with min.per.group=1L, and the reorganize. But I got error as there are NAs in the merged data.
Thanks!
