Entering edit mode
9.1 years ago
rdbcasillas11
▴
10
Does anyone know how to convert MEDIP-seq RMS(raw methylation score) values into AMS? I looked into MEDIPS package but the documentation isn't clear on how to convert it. I have MEDIP RMS values from a patient and would like to convert them into 0-1000 range of AMS using MEDIPS if possible.
Thank you.
Thanks Devon. The tutorial in the pdf you referred to is actually difficult to follow because the extdata directory(inside MEDIPS package) that they use(and say comes along with the package) is empty and has no files to work with. Also, the MEDIPS.methylProfiling function referred to is deprecated.
That's annoying. You might be best off contacting the authors. MEDIPS isn't a particularly heavily used package, since the technique is falling out of vogue.
Is there any other package that you would recommend for converting RMS into AMS? Or is that something only MEDIPS does?
Maybe Batman? It's been a long time since I've used either of those, but it's targeted at the same questions so perhaps it can convert to absolute methylation. Note that you often need a calibration dataset regardless of the package (simply so you can appropriately determine the background coverage biases).
If you follow the newer MEDIPS tutorial from April 2015 (http://www.bioconductor.org/packages/release/bioc/vignettes/MEDIPS/inst/doc/MEDIPS.pdf) it explains that you have to download the tutorial data by calling
> source("http://bioconductor.org/biocLite.R")
> biocLite("MEDIPSData")
The tutorial data should appear in your extdata directory after calling those commands!
Yeah, I was able to find extdata earlier using the steps you mentioned. My original question of converting RMS into AMS seems to be not so relevant anymore as the MEDIPS author himself advises against doing that and considers RMS values to be good enough for comparing 'MSets'.