I am trying to test each CpG cluster for the presence of a differentially methylated location using the testClusters
function, however I am not getting any result unless I use FDR.cluster = 1
. Anything below 1, I get an error like this:
Error in testClusters(locCor, FDR.cluster = 0.1) :
No CpG clusters rejected.
If I continue the procedure using FDR.cluster = 1
, then when I get to the trimClusters
step
clusters.trimmed <- trimClusters(clusters.rej, FDR.loc = 0.05)
no matter what number I use for FDR.loc
, I get the error message:
Error in integrate(integrand, lower = z.li[loc], upper = Inf) :
non-finite function value
I am worried that the number I am using for FDR.cluster
might be too high, and is what's causing the error in the trimCluster
step, but I am not sure. Does anyone have any idea if this is an ok number, or if not, why it might be so high, and why I might be getting the error?
This is my session info: R version 4.0.0 (2020-04-24) Platform: x86_64-apple-darwin17.0 (64-bit) Running under: macOS Catalina 10.15.4
Matrix products: default BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib
locale: [1] enCA.UTF-8/enCA.UTF-8/enCA.UTF-8/C/enCA.UTF-8/en_CA.UTF-8
attached base packages:
[1] parallel stats4 stats graphics grDevices utils
[7] datasets methods base
other attached packages:
[1] BiSeq1.28.0 Formula1.2-3
[3] SummarizedExperiment1.18.1 DelayedArray0.14.0
[5] matrixStats0.56.0 Biobase2.48.0
[7] GenomicRanges1.40.0 GenomeInfoDb1.24.0
[9] IRanges2.22.1 S4Vectors0.26.0
[11] BiocGenerics_0.34.0
loaded via a namespace (and not attached):
[1] Rcpp1.0.4.6 compiler4.0.0
[3] XVector0.28.0 bitops1.0-6
[5] tools4.0.0 zlibbioc1.34.0
[7] digest0.6.25 bit1.1-15.2
[9] memoise1.1.0 RSQLite2.2.0
[11] annotate1.66.0 lattice0.20-41
[13] rlang0.4.6 Matrix1.2-18
[15] DBI1.1.0 rstudioapi0.11
[17] GenomeInfoDbData1.2.3 rtracklayer1.48.0
[19] vctrs0.2.4 Biostrings2.56.0
[21] bit640.9-7 globaltest5.42.0
[23] lmtest0.9-37 grid4.0.0
[25] nnet7.3-14 flexmix2.3-15
[27] AnnotationDbi1.50.0 survival3.1-12
[29] XML3.99-0.3 BiocParallel1.22.0
[31] lokern1.1-8 blob1.2.1
[33] splines4.0.0 Rsamtools2.4.0
[35] modeltools0.2-23 sfsmisc1.1-6
[37] GenomicAlignments1.24.0 xtable1.8-4
[39] betareg3.1-3 sandwich2.5-1
[41] RCurl1.98-1.2 crayon1.3.4
[43] zoo_1.8-8
If you don't get any significant clusters at a reasonable FDR threshold (usually 5%, or 25% if you are really desperate) then there are no differentially methylated regions in your data. You may not have enough replicates per group?