Difference between bismark2bedGraph and methylkit CpG calling

[dude.d...@gmail.com]

Hi,

  I am wondering why there is a difference in the number of CpG methylation calling when I used the build-in function of bismark-bismark2bedGraph and methylkit.

	Bismark2bedGraph
	treatment_rep1	treatment_rep2	Control
10 read count filter(done by R)	5023385	4906324	4821601
common position (done by R)	3663137	3663137	3663137

	Methykit
	treatment_rep1	treatment_rep2	Control
10 read count filter	5468215	5402572	5327155
common position	4369283	4369283	4369283

Furthermore,

I also used the calculateDiffMeth. Although the q-value of some sites are extremely low, when I convert it the biswig and visualize on UCSC, those sites are only be differentially methylated in one of the replicates, not the both. Is it due to small sample size? I appreciate any suggestions and I hope someone can help me.

Thanks.

[Altuna Akalin]

I can't comment on what bismark2bedgraph is doing. You have to ask them or look at their code to see what kind of filtering they apply. We simply stack up the reads and use bismark methylation calls to call percent methylation on a base.

We need have at least 10X coverage defaul and fastq quality score for a base should be >20 phred score default. Maybe they are doing extra filtering. 

 

Furthermore,

I also used the calculateDiffMeth. Although the q-value of some sites are extremely low, when I convert it the biswig and visualize on UCSC, those sites are only be differentially methylated in one of the replicates, not the both. Is it due to small sample size? I appreciate any suggestions and I hope someone can help me.

I think there is no point to try to use logistic regression based test when you don't have replicates for control. i would use pool() to pool the tests or compare test1 vs control and test2 vs control and take the consensus. 

Best,

Altuna

 

Thanks.

--
You received this message because you are subscribed to the Google Groups "methylkit_discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to methylkit_discus...@googlegroups.com.
To post to this group, send email to methylkit_...@googlegroups.com.
Visit this group at https://groups.google.com/group/methylkit_discussion.
For more options, visit https://groups.google.com/d/optout.