Question: QSEA: Need help on creating ROI for whole genome MeDIPSeq data
gravatar for geekyvt
24 days ago by
geekyvt0 wrote:

I am analyzing 4 MeDIP Seq samples. I wanted to know,  how do I create ROI. As of now, I do not have any shortlisted regions, I want to look at entire human genome. How to get it done. 

Thanks in advance for any help.


ADD COMMENTlink modified 24 days ago by Matthias Lienhard140 • written 24 days ago by geekyvt0
gravatar for Matthias Lienhard
24 days ago by
Max Planck Institute for molecular Genetics, Berlin, Germany
Matthias Lienhard140 wrote:

Hi Vivek,

in qsea, makeTable offers two possibilities to look at regions of interest (ROIs):

The "ROIs" parameter takes a GRange object selects all windows that overlap with these Genomic Ranges. Windows overlapping with more than one ROI are represented several times.

The "annotation" parameter takes a list  of GRange objects (e.g. CpG islands, promoters, TFBS) and annotates the resulting table with the information from the GRanges. For each element of the list one column will be added, containing information from the metadata columns (mcols) of the overlapping GRanges. This is described in step 3.6 "Annotating, Exploring and Exporting Results" of the tutorial.

Additionally, you can compute the number of overlapping DMRs per class of ROIs using the regionStats() function, such as outlined in the tutorial. These numbers can then be used for overrepresentation analysis.

You can create GRanges from tables, for example downloaded from the UCSC table browser for the genome build you are using. For details look at the vignettes of "makeGRangesFromDataFrame()" and "GRanges()"  functions of the GenomicRanges package.

Best, Matthias

ADD COMMENTlink written 24 days ago by Matthias Lienhard140

Dear Matthias,

Thank you for explaining this in detail. I still stuck with one place, ROIs is used in the below step,

result=makeTable(qseaSet, glm=qseaGLM, groupMeans=getSampleGroups(qseaSet), keep=sig, annotation=ROIs, norm_method="beta")

I guess this has been taken directly from data(annotation, package="MEDIPSData"). Is there a way to create it for entire genome within QSEA..? Or maybe directly annotated every region identified through maketable command.

Thanks in advance,



ADD REPLYlink written 24 days ago by geekyvt0

Dear Vivek,

retrieving annotation is not within the scope of qsea. You are right, for the tutorial I provide a example for the analyzed chromosomes. You can create these objects yourself, by downloading tables from the UCSC table browser, import them in R (e.g. by read.table()) and then create a GRange object from it, using "makeGRangesFromDataFrame()" or "GRanges()". Alternativly, you can have a look at the annotation packages in bioconductor, such as outlined in this presentation:

ADD REPLYlink written 24 days ago by Matthias Lienhard140
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.2.0
Traffic: 201 users visited in the last hour