Are genomic context categories mutually exclusive?

[IRECG]

Hello,
I am using Goldmine to annotate my DMRs from methylaction, and reading your paper published on Cell Reports I have a question. In that work you say: "Categories were made mutually exclusive in cases of multiple overlaps by using the priority: promoter, 30 end, exon, intron, intergenic." My question is if this is the default procedure that Goldmine uses.
If I understand it well a DMR is not going to be annotated in more than one of the genomic context options, but it can be a promoter and a CpG island, for instance.

Thanks,
Irene

[jeffbhasin]

Hi Irene,
Yes the context categories are mutually exclusive, using the priority order. In case of multiple annotations, only the one with highest rank in the priority order is reported. However, Goldmine provides a lot of details so the other overlaps can be enumerated from the other columns. This is only for the gene model categories - for features like CpG island are not reported in the call column. Each feature has their own column and aren't part of the priority order. Using a threshold on the percent overlap reported in this column, one can extract lists of DMRs that overlap with given feature set(s).

Jeff

[IRECG]

Hi,
So the one that I see on the "call" column is the one that has been chosen using the priority order (in case there are more than one annotations), right?
Thanks again,
Irene

[jeffbhasin]

Yes the call column is generated using the priority order.