Entering edit mode
12 days ago
Cassandra
•
0
Hello all,
I have a WGBS dataset with 1 control and 6 diseased samples. Has anyone else encountered this structure and how would you recommend analyzing it? I see methylKit supports a 1-1 ratio (although it is optimized for RRBS) but I have not found anything about 1 control to multiple experimental samples.
Thanks!
For WGBS analysis with multiple samples you can solve it by one by one. Although you have only 7 samples. To analyze the WGBS file you can use Linux operating system but the RAM should be minimum 60 GB and storage 2TB and CPU 16 core 32 thread for better performance.
What is your analysis goal? Differential methylation? edgeR has a workflow (see its user guide) where 6 vs 1 would be supported. That is obviously limited by the fact that you assume the 1 control sample was representative, but at least it does proper statistics given the replication int he case group.