How to convert GenomicRatioSet to RGChannelSet in minfi::compartments()
1
0
Entering edit mode
stewart999 ▴ 10
@stewart999-11769
Last seen 13 months ago

I am trying to estimate A/B compartments using minfi's compartments function. The data I have is from whole blood samples, so I want to correct for cellular heterogeneity first. I have done this using minfi's estimateCellCounts function with the option returnAll = TRUE.

This returns data corrected for cell counts in the form of a GenomicRatioSet. I can see from Fortin and Hansen's (2016) paper describing the use of methylation data to predict A/B compartments that it is recommended to normalise the data first using functional normalisation.

The minfi function preprocessFunNorm takes an RGChannelSet as input. Is there a way that I can convert the GenomicRatioSet created after correcting for cellular heterogeneity into an RGChannelSet to use in preprocessFunNorm. If not, does anyone have any recommendations for other approaches to pre-processing of whole blood methylation data prior to estimating A/B compartments?

Example code:

x <- minfi::estimateCellCounts(RGset, returnAll=TRUE, verbose=TRUE)

x is a list of 3 objects:

  1. counts        : num
  2. compTable     :'data.frame'
  3. normalizedData:Formal class 'GenomicRatioSet'

 

minfi estimateCellCounts methylation • 948 views
ADD COMMENT
0
Entering edit mode
@kasper-daniel-hansen-2979
Last seen 3 months ago
United States
The output from estimateCellCounts is not corrected for cell type composition. The `normalizedData` returned by the function is probably not something you care about. It is your input data after being normalized together with the reference data. To correct for cell type composition you need to use the estimated cell type proportions. We don't know how to do this for the AB compartment estimation, and I don't know for sure whether it works on blood data. In our paper we show an example of a dataset collected in whole blood, where the estimation fails, but we are not sure whether this is general to all blood datasets or whether it is something specific to that dataset. You can run the estimation yourself and do the QC we recommend to assess this. Best, Kasper On Thu, May 18, 2017 at 4:50 AM, stewart999 [bioc] <noreply@bioconductor.org> wrote: > Activity on a post you are following on support.bioconductor.org > > User stewart999 <https: support.bioconductor.org="" u="" 11769=""/> wrote Question: > How to convert GenomicRatioSet to RGChannelSet in minfi::compartments() > <https: support.bioconductor.org="" p="" 96080=""/>: > > I am trying to estimate A/B compartments using minfi's compartments > function. The data I have is from whole blood samples, so I want to correct > for cellular heterogeneity first. I have done this using minfi's > estimateCellCounts function with the option returnAll = TRUE. > > This returns data corrected for cell counts in the form of a > GenomicRatioSet. I can see from Fortin and Hansen's (2016) paper describing > the use of methylation data to predict A/B compartments that it is > recommended to normalise the data first using functional normalisation. > > The minfi function preprocessFunNorm takes an RGChannelSet as input. Is > there a way that I can convert the GenomicRatioSet created after correcting > for cellular heterogeneity into an RGChannelSet to use in > preprocessFunNorm. If not, does anyone have any recommendations for other > approaches to pre-processing of whole blood methylation data prior to > estimating A/B compartments? > > Example code: > > x <- minfi::estimateCellCounts(RGset, returnAll=TRUE, verbose=TRUE) > > x is a list of 3 objects: > > 1. counts : num > 2. compTable :'data.frame' > 3. normalizedData:Formal class 'GenomicRatioSet' > > > > ------------------------------ > > Post tags: minfi, estimateCellCounts, methylation > > You may reply via email or visit How to convert GenomicRatioSet to RGChannelSet in minfi::compartments() >
ADD COMMENT

Login before adding your answer.

Traffic: 199 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6