I am attempting to run DEXSeq on an RNASeq data set to determine if there is differential splicing due to treatment condition. However, during the estimate dispersions step I get the following error: simpleLoess(y, x, w, span, degree, parametric, drop.square, normalize, : NA/NaN/Inf in foreign function call (arg 1) What would cause such an error given the experimental design and provided design formulas?

My experimental set up has 23 patients, and each patient has two conditions - time 0 (t0) and time 1 (t1).The sample metadata looks like the following:

PatientID    SampleID    Condition

p1              p1-t0           t0

p1              p1-t1           t1

p2              p2-t0           t0

p2              p2-t1           t1

…              …               ...

p23              p23-t0           t0

p23              p23-t1           t1

Due to patient variability, I would like to block by PatientID. Following the DEXSeq vignette example, I set up my design formula and null model as the following:

> design.dex <- ~ sample + exon + PatientID:exon + Condition:exon
> null.model <- ~ sample + exon + PatientID:exon

I am using the wrapper function to complete the entire analysis with the following command: 

> output <- DEXSeq(dexSeqdataset,fullModel = design.dex, reducedModel = null.model, BPPARAM = MulticoreParam(workers=8),fitExpToVar="Condition",quiet=FALSE)

> sessionInfo()
R version 3.1.0 (2014-04-10)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods  
[8] base     

other attached packages:
 [1] DEXSeq_1.10.6             BiocParallel_0.6.1       
 [3] DESeq2_1.4.5              RcppArmadillo_0.4.300.8.0
 [5] Rcpp_0.11.2               GenomicRanges_1.16.3     
 [7] GenomeInfoDb_1.0.2        IRanges_1.22.9           
 [9] Biobase_2.24.0            BiocGenerics_0.10.0      
[11] BiocInstaller_1.14.2     

loaded via a namespace (and not attached):
 [1] annotate_1.42.0      AnnotationDbi_1.26.0 BatchJobs_1.2       
 [4] BBmisc_1.7           biomaRt_2.20.0       Biostrings_2.32.0   
 [7] bitops_1.0-6         brew_1.0-6           checkmate_1.0       
[10] codetools_0.2-8      DBI_0.2-7            digest_0.6.4        
[13] fail_1.2             foreach_1.4.2        genefilter_1.46.1   
[16] geneplotter_1.42.0   grid_3.1.0           hwriter_1.3         
[19] iterators_1.0.7      lattice_0.20-29      locfit_1.5-9.1      
[22] plyr_1.8.1           RColorBrewer_1.0-5   RCurl_1.95-4.1      
[25] Rsamtools_1.16.1     RSQLite_0.11.4       sendmailR_1.1-2     
[28] splines_3.1.0        statmod_1.4.20       stats4_3.1.0        
[31] stringr_0.6.2        survival_2.37-7      tools_3.1.0         
[34] XML_3.98-1.1         xtable_1.7-3         XVector_0.4.0       
[37] zlibbioc_1.10.0     

Hi cfreije,

This issue has to do with your big number of samples and should be solved in the development version of DEXSeq! Could you try to update to these and give it a try?

Alejandro