kissDE error during 'diffExpressedVariants' step: "Error in data.frame..."

Question

[TrentGenomics]

Hi,

I'm running kissDE on the results obtained from KisSplice. I used 4 paired-end samples as input for KisSplice, with each of the 8 FQ files totaling ~ 13 gigabytes each.

Using R on a high performance computer cluster, kissDE ran for ~12 days, before reaching this error during the 'diffExpressedVariants' step:

res<-diffExpressedVariants(snp, species, pvalue=1)

[1] "Pre-processing the data..."

[1] "Trying to fit models on data..."

[1] "Searching for best model and computing pvalues..."

[1] "Error in data.frame(..., check-names = FALSE): arguments imply differing number of rows: 537429

Is any one able to provide more information on this error? Any info would be appreciated.

Cheers,

Answer 1

[leandro.ishi.lima]

Dear TrentGenomics,

In order to debug better this error (which we have not seen before), could you please:

1) Post the command line you used for KisSplice and kissDE?

2) What your "species" vector contain?

3) Print the head of your "snp" data.frame?

4) Did you do the quality control step of kissDE?

We guess that the reason for this error could be because the same counts parameters were not used when running both KisSplice and kissDE, or because the conditions vector is not suited for your analyses. With your answers to the previous questions, we can understand better what is happening.

Thank you very much.