I'm not super familiar with different data formats for methylation data, but I have a raw data matrix with the beta values (B = methylation signal / (methylation signal + unmethylated signal)) in matrix form, where the rows are the CpG loci IDs (Illumina 450K is the platform) and the columns are the sample IDs.
I'm trying to load my data into RnBeads package with the command rnb.run.analysis(dir.reports=reports.dir, sample.sheet=sample.annotation, data.dir=raw.dir, datatype=??). I'm not sure which data type a beta-matrix is. Can anyone help me out? The vignette is not clear at all to me! I get the furthest when I do the below, but I get the error message: Error in rnb.region.types.for.analysis("hg19"): invalid value for rnb.set
dirs <- list(data="~/Downloads/450K/raw.data/",
annotation="~/Downloads/450K/raw.data/sample_annotation.txt",
analysis="~/Downloads/450K/results/",
reports="~/Downloads/450K/results/reports")
rnb.getOption("import.table.separator")
rnb.options(filtering.sex.chromosomes.removal=TRUE,
import.table.separator="\t")
result <- rnb.run.analysis(dir.reports=dirs$reports,
data.source=dirs$data,
data.type="infinium.data.dir")
Secondly, what happens if you have no sample annotation file? I created a 2 column file based on my intuition from the sample IDs, where the first column are the sample IDs and the second column identifies unique people the samples are from (it's just the number from the sample IDs). I have no other information about the samples regarding well #, sex, etc. Is this a problem for RnBeads' pipeline?
Thanks for any insights!
