crlmm package in R - processing the SNP6 copy number data error No arrays above quality threshold!

Question

[misterie]

Hi,

I am trying to analysis SNP6.0 copy number data. For this purpose I am using the crlmm package in R. The first steps look like;

cnSet <- constructAffyCNSet(sinfo$celFiles, batch=rep(1, nrow(sinfo)),cdfName="genomewidesnp6",
                              genome="hg19")
cnrmaAffy(cnSet)
snprmaAffy(cnSet)
validCEL(sinfo$celFiles)
genotypeAffy(cnSet)

All the lines pass properly except the genotypeAffy function that returns:

Error in crlmmGT2(A = A(cnSet), B = B(cnSet), SNR = cnSet$SNR, mixtureParams = cnSet@mixtureParams, : No arrays above quality threshold!

Do you have any idea, how can I resolve this issue?

Answer 1

[Kevin Blighe]

Check some of the parameters that can be passed to genotypeAffy(), namely, these: https://www.rdocumentation.org/packages/crlmm/versions/1.30.0/topics/genotypeAffy

How many samples have you got? What is the source tissue? If it is something like FFPE DNA, then many probes will fail.

[misterie]

Hi, thank you. There are 11 samples.

[Kevin Blighe]

That is a low number. Take a look at the parameters that can be passed to the function.

[misterie]

Thank you. I checked these parameters. Now I increased the number of samples to 80 and I get just a warning:

Warning message:
In crlmmGT2(A = A(cnSet), B = B(cnSet), SNR = cnSet$SNR, mixtureParams = cnSet@mixtureParams,  :
  Recalibration not possible. Possible cause: small sample size.

is it still a problem?

[Kevin Blighe]

Yes and no... the issue is now the parameter recallRegMin=1000. I would reduce that progressively until the warning diappears; however, if you get below, e.g., 500, then start to worry...