Skip to content

Spatial deconvolution & multiple negative probes #54

@kristakokki9-cmd

Description

@kristakokki9-cmd

Hi (I think @ningbioinfo ),

I've been attempting to run Spatial Decon for WTA data, which has multiple negative probes. (I've read the discussions / issues here regarding them, but am still at loss).

Here's the code up to spatial decon:

spe2 <- readGeoMx(countFile2, sampleAnnoFile, featureAnnoFile2, rmNegProbe = F)
spe2 <- addPerROIQC(spe2, rm_genes = TRUE)
qc <- colData(spe2)$AOINucleiCount > 100
spe2 <- findNCGs(spe2, batch_name = "SlideName", top_n = 300)
spe_ruv2 <- geomxBatchCorrection(spe2, factors = "SegmentLabel", 
                                 NCGs = metadata(spe)$NCGs, k = 4)

I have 139 negative probes, named NegProbe-WTX_1, NegProbe-WTX_2, NegProbe-WTX_3 and so on. The names are unique.

When I attempt to run prepareSpatialDecon, I get an error saying that NegProbe-WTX must be included in the dataset. If I understand correctly, it would require exact name match, "NegProbe_WTX", which I do not have - admittedly, it does work if I instead but "NegProbe-WTX1", but I doubt that's the correct play here.

##So this works, in theory
spd=prepareSpatialDecon(spe_ruv2, negProbeName = "NegProbe-WTX_1",
                        pool = nrow(spe_ruv2))

So how do I run prepareSpatialDecon so that it handles multiple negative probes?

Secondly, regarding the same function; I understand argument "pool" requires a number. Is that number supposed to be number of negative probes in the dataset, number of genes or number of negative probes + genes (aka number of rows). I did test with all the values, and resulting data frames are exact matches, through.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions