delta Beta as output feature

[bopohdr]

Hi!

I was struggling with epimutation identification (e.g. ramr package on the Barbosa dataset did not provide much true positive, validated calls http://www.bioconductor.org/packages/release/bioc/html/ramr.html ), so thanks a lot you developing such useful tool (that actually works!).

I know it is still in active development, but one thing I miss in the output is - mean beta value for the region in controls and in the case (and delta beta), as this really helps the interpretation of the epimutations and understanding their potential effect (e.g. epimutation with delta beta 0.7 vs. 0.15 would likely be interpreted a bit differently). These values are also usually reported in all papers (including supplements of Barbosa et al.). Therefore, it would be great and very helpful if you could implement it as well.

Thanks!

Best,
Dmitrijs

[isglobal-brge]

Hi Dmitrijs,

first of all, thank you very much for your interest in our work! We also appreciate your input! Actually, we are going to implement what you are indicating.

Just one comment. Could you, please, send us any reference of papers using "delta beta" values and how they use them in practice? We would like to comment on that in our vignette and our paper (we are almost finishing)

Thanks a lot!

Juan

[bopohdr]

Hi !

Great!

1. Actually in the paper of Barbosa et al., 2018 in the Supplementary Data 5 mean Beta values in cases vs. parents vs. controls are used to show the de novo status and compare with controls.
2. the "ramr" package is also outputting the delta mean beta value
3. https://doi.org/10.1016/j.ajhg.2016.06.032 this study (also by Andry Sharp) does not use delta Beta, but mean betas of the DMR

Hope this helps!

Best,
Dmitrijs

[yocra3]

This issue is partially solved. The methods now returns the delta_beta but cannot manage having a NA in any of the reference samples.