Question on import methylation intensity

[yeswzc]

Hi, Do anyone know how to import methylation intensity signal matrix as a RnBeads or methylumi object? I have a matrix with columns as samples with their "Methylated Signal", "Unmethylated Signal" and "Detection Pval" downloaded from GEO database. The raw idat files were not available so I have to start from the intensity signals.
Is there a function can be used to import the data so that I can use other functions in methylumi to normalize the data?

Thank you!

[schmic05]

Hi. I think the constructor for RnBeadRawSet is what you are searching for. Once you creating this object using the M and U arguments, you can use all of RnBeads' functionality on this object. Check out the documentation using ?RnBeadRawSet

[yeswzc]

Thank you! I found the usage of RnBeadRawSet requires "pheno" matrix as input. Do you know what is "pheno" and how can I creat the "pheno" matrix?

RnBeadRawSet(pheno, probes, M, U, M0 = NULL, U0 = NULL,
bead.counts.M = NULL, bead.counts.U = NULL, p.values = NULL,
qc = NULL, platform = "450k", beta.offset = 100,
summarize.bead.counts = TRUE, summarize.regions = TRUE,
region.types = rnb.region.types.for.analysis("hg19"),
useff = rnb.getOption("disk.dump.big.matrices"), ffcleanup = FALSE)

[yeswzc]

Thanks. I found an example of the pheno matrix.

Sample_ID Sentrix_ID Sentrix_Position   Sample_Plate Sample_Well Sample_Group  Cell_Line Passage_No Treatment           barcode
 hES_HUES13_p47 5815381013           R03C01 WG0001341-MSA4         C01     hESC hES_HUES13         47      <NA> 5815381013_R03C01

Some column can be ignored.

Best.

[yeswzc]

When imported by RnBeadRawSet(), do you know how can I used method in RnBeads to normalize the data? I get error below:

> rnb.execute.normalization(obj2)
Object of class RnBeadRawSet
     696 samples
  485577 probes
        of which: 482421 CpG, 3091 CpH, and 65 rs
Region types:
          138358 regions of type tiling
           31033 regions of type genes
           31195 regions of type promoters
           26662 regions of type cpgislands
Intensity information is present
Detection p-values are present
Bead counts are absent
Quality control information is absent
Summary of normalization procedures:
        The methylation data was not normalized.
        No background correction was performed.

Thanks!

[schmic05]

I don't see an error in your above output. The message that you see is the summary of the new object that has been created. You will have to show me the actual error message that is printed during the normalization process. Perhaps it helps to change the normalization method using rnb.options(normalization.method="wm.dasen").

[yeswzc]

Sorry. My fault. There is no error. Is there a way to background correct the RnBeadRawSet and normalize it using functions in methylumi method? Seems the data do not have the required information?

Thank you!

[schmic05]

Yes, RnBeads supports some of the normalization methods available in methylumi, see the documentation of ?rnb.execute.normalization for further information. As the names suggest, you will also need to have intensity information about the background control probes on the Illumina arrays to perform background correction.

[yeswzc]

Understand now. Thanks you!