Question

Conumee for EPIC and EPICv2 copy number

0

Entering edit mode

Adam • 0

@778deb0d

Last seen 5 hours ago

United States

I am attempting to extract copy number from EPIC and EPICv2 arrays.

I have tried some things but run into the same issues as reported here, here, and here.

CNV.fit() -> "query intensities not given for all probes"

CNV.segment() -> "Error in if (min(weights) <= 0) stop("all weights should be positive") : missing value where TRUE/FALSE needed"

I am seeing in ?CNV.create_anno that array_type overlap.2 would convenient but I haven't gotten it to work.

Is array_type = "overlap.2" available?

If not, has anyone been able to handle EPIC and EPICv2 together.

> ?CNV.create_anno
...
array_type: character. One of '450k', 'EPIC', 'EPICv2', 'mouse'. When
          analyzing data from multiple array types, choose between
          'overlap.1' for 450k and EPIC, 'overlap.2' for EPIC and
          EPICv2 or 'overlap.3' for 450k, EPIC, EPICv2.
...

> anno <- CNV.create_anno(array_type = "overlap.2", exclude_regions = exclude_regions, detail_regions = detail_regions)
Error in CNV.create_anno(array_type = "overlap.2", exclude_regions = exclude_regions,  : 
  array_type must be one/multiple of 450k, EPIC, EPICv2, or mouse

> packageVersion("conumee2")
[1] '2.1'

conumee CopyNumberVariation MethylationArrayData • 64 views

ADD COMMENT • link 22 hours ago • updated 6 hours ago Adam • 0

0

Entering edit mode

As described on the github the way to do this is using array_type = c("EPIC", "EPICv2")

This is working for me now so I'll leave it here. I have all EPIC controls, but this should work with both.

library(conumee2)
library(sesame)

sdfs.q <- openSesame(".", prep="QCDPB", func=NULL)
sdfs.c <- openSesame("controls", prep="QCDPB", func=NULL)

data(exclude_regions)
data(detail_regions)
anno <- CNV.create_anno(array_type = c("EPIC", "EPICv2"), exclude_regions = exclude_regions, detail_regions = detail_regions) 

sdfs.q <- mLiftOver(sdfs.q, "EPIC")
#sdfs.c <- mLiftOver(sdfs.c, "EPIC")
sdfs.q <- lapply(sdfs.q, function(sdf) sdf[sdf$Probe_ID %in% names(anno@probes), ])
#sdfs.c <- lapply(sdfs.c, function(sdf) sdf[sdf$Probe_ID %in% names(anno@probes), ])

data.q <- CNV.load(do.call(cbind, lapply(sdfs.q, totalIntensities)))
data.c <- CNV.load(do.call(cbind, lapply(sdfs.c, totalIntensities)))

x <- CNV.fit(data.q, data.c, anno)
x <- CNV.bin(x)
x <- CNV.detail(x)
x <- CNV.segment(x)

ADD REPLY • link 7 hours ago • updated 6 hours ago Adam • 0