is it necessary to consider strand direction when mapping CpG sites to regions of genes?

Question

[liu4gre]

Dear all,

I have CpG methylation from 450k and RRBS. Now I am trying to map these CpGs to genes' different regions, e.g. around TSS. My concerns is if the direction (+ or -) should be considered here. For example, probes from 450k have directions given by the annotation file, and RRBS also gives the directions of reads. So should I map them into the same-direction genes?

Thanks.

Answer 1

[Devon Ryan]

Individual Cs will have a strand and about this is present in the RRBS reads. So that much is worthwhile considering. Now the perhaps your question is if you have a gene on the + strand whether you should only look at methylation on the + strand. The answer to that is no, you should look at both strands' methylation.