Missing gene in the DROP OUTRIDER_results_all.Rds

[lyj95618]

Hi,

I tried to use the expression analysis module in the DROP pipeline(v1.6.1). When the pipeline finished, I don't see my gene of interest in the OUTRIDER_results_all.Rds. I checked the ods_fitted.Rds, ods_unfitted.Rds, ods.Rds using OUTRIDER_results_all@assays@data$counts[mygeneID,] and was able to see my gene here. All the samples have at least 500 counts for this gene. Why is this gene not in the final output?

my config setting

aberrantExpression:
    run: true
    groups:
        - FIBRO
    fpkmCutoff: 1
    implementation: autoencoder
    padjCutoff: 0.05
    zScoreCutoff: 0
    genesToTest: null
    maxTestedDimensionProportion: 3
    yieldSize: 2000000
    useOHTtoObtainQ: true

Thanks for the help!
Laur

[vyepez88]

Hi, can you confirm that when loading the ods_fitted.Rds or ods.Rds, you are able to run this:
ods[geneID, ]
?
If not, it's probably filtered out. For that, check the following on the ods_unfitted.Rds
sort(fpkm(ods_unfitted)[geneID,])

[lyj95618]

Hi,

Here are the outputs. I got outputs from ods_fitted.Rds, ods.Rds, and ods_unfitted.Rds

fit_ods <- readRDS('./ods_fitted.Rds')
fit_ods['ENSG00000187240',]

class: OutriderDataSet
class: RangedSummarizedExperiment 
dim: 1 53 
metadata(10): version encDimTable ... loss convList
assays(6): counts normalizationFactors ... l2fc zScore
rownames(1): ENSG00000187240
rowData names(9): passedFilter basepairs ... FitDMessage
  expectedLogGeomMean
colnames(53): sample1 sample2 ... sample52
  sample53
colData names(21): sampleID RNA_ID ... sizeFactor thetaCorrection

my_gene_fpkm <- sort(fpkm(fit_ods)['ENSG00000187240',]) # this give a list of number I think its the fpkm of my gene of interest?
sum(my_gene_fpkm >= 1) # 42 of 53 of the samples have fpkm >= 1
42

unfit_ods <- readRDS('./ods_unfitted.Rds')
unfit_ods['ENSG00000187240',]

class: OutriderDataSet
class: RangedSummarizedExperiment 
dim: 1 53 
metadata(1): version
assays(1): counts
rownames(1): ENSG00000187240
rowData names(3): passedFilter basepairs counted1sample
colnames(53): sample1 sample2 ... sample52
  sample53
colData names(19): sampleID RNA_ID ... expressedGenesRank isExternal

sort(fpkm(unfit_ods)['ENSG00000187240',])#give me a list of number again

I found that fpkm(unfit_ods)['ENSG00000187240',] and fpkm(fit_ods)['ENSG00000187240',] gave slightly different values for the same sample.

Thanks for the help!
Laur

[lyj95618]

Hi I am able to find my gene of interest now. I think it was just my mistake. Thanks! I will close the issue!