Closed ChengYunazhi closed 6 years ago
hi @ChengYunazhi Defiant was designed to identify differentially methylated regions. However, you can run defiant with the "-E" option to get information about every CpG/CpN site and compare on a whole genebody level as you want. This will give you a p-value and percent difference for every CpG/CpN site in a file that starts with "nucleotide". You can compare different regions this way. You can paste attach the output file here, if you like, and I can help.
Thanks David, I'll try this "-E" option.
Hi, @hhg7
Is there a statistical method that gives just one p-value for differential methylation of ROIs? I used defiant to get a p-value for ever CpG in ROIs, but how to assess whether there is a significant difference in mean methylation level between the two sample or two groups in these ROIs? A similar analysis can be done in the MEDIP-seq data, I don’t know if this method is feasible in WGBS data?
Thanks
Yuanzhi
hi @ChengYunazhi I will write a script to do what you ask, it might be a few days. It will process the output file from the "-E" option.
Thanks for your kind help, @hhg7 . Is it possible to have the method address datasets with or without biological replicates?
hi @ChengYunazhi yes, if there are no biological replicates the program uses Fisher's exact test instead of Welch's t-test.
@ChengYunazhi I've uploaded the program roi.c onto the main repository. Please let me know how it runs. I look forward to your response.
Thanks @hhg7 , I will run it and give you feedback !
@ChengYunazhi did the roi program do what you intended? please let me know, I'd like to close the issue.
Hi,
Thanks!
Yuanzhi