Dear List

I am trying to customise the cdc file for the Affymetrix mirna20 chip (mirna20cdf). Following the directions in the vignette ‘Modifying existing CDF environments to make
alternative CDF environments’ in the altcdfenvs package I have (data_in is from a ReadAffy call):

library(altcdfenvs)
library(mirna20cdf)

print(data_in)

AffyBatch object
size of arrays=478x478 features (132 kb)
cdf=mirna20cdf (20706 affyids)
number of samples=276
number of genes=20706
annotation=mirna20
notes=

reduced_cdf <- wrapCdfEnvAffy(mirna20cdf, 478, 478, 'mirna20cdf')

However this fails with the following error:

Error in if (any(i <= 0)) stop("Indices must start at 0 or 1 (please refer to the help file) !") :
  missing value where TRUE/FALSE needed

Similarly the getCdfEnvAffy command also fails.

getCdfEnvAffy(data_in)
Error in if (any(i <= 0)) stop("Indices must start at 0 or 1 (please refer to the help file) !") :
  missing value where TRUE/FALSE needed

If anyone can help with this I’d be grateful.

Best,

Iain

> sessionInfo()
R version 3.2.0 (2015-04-16)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X 10.10.4 (Yosemite)

locale:
[1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8

attached base packages:
[1] stats4    parallel  stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] altcdfenvs_2.30.0     hypergraph_1.40.0     graph_1.46.0          Biostrings_2.36.1    
 [5] XVector_0.8.0         IRanges_2.2.5         S4Vectors_0.6.2       makecdfenv_1.44.0    
 [9] affyio_1.36.0         mirna20cdf_2.16.0     data.table_1.9.4      ggplot2_1.0.1        
[13] dplyr_0.4.2           reshape2_1.4.1        class_7.3-13          affyPLM_1.44.0       
[17] preprocessCore_1.30.0 gcrma_2.40.0          vsn_3.36.0            limma_3.24.14        
[21] genefilter_1.50.0     affy_1.46.1           Biobase_2.28.0        BiocGenerics_0.14.0  
[25] RColorBrewer_1.1-2    BiocInstaller_1.18.4

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.0          GenomeInfoDb_1.4.1   plyr_1.8.3           tools_3.2.0         
 [5] zlibbioc_1.14.0      digest_0.6.8         gtable_0.1.2         annotate_1.46.1     
 [9] RSQLite_1.0.0        lattice_0.20-33      DBI_0.3.1            proto_0.3-10        
[13] stringr_1.0.0        grid_3.2.0           R6_2.1.0             AnnotationDbi_1.30.1
[17] XML_3.98-1.3         survival_2.38-3      magrittr_1.5         MASS_7.3-43         
[21] scales_0.2.5         splines_3.2.0        assertthat_0.1       colorspace_1.2-6    
[25] xtable_1.7-4         labeling_0.3         stringi_0.5-5        munsell_0.4.2       
[29] chron_2.3-47   

The error comes from the fact that the miRNA 2.0 array is technically a PM/MM array, but in general is not. In other words, all of the Affymetrix control probesets have MM probes, but none of the probesets that measure miRNAs have MM probes.

This causes problems for all the non-control probesets, because internally wrapCdfEnvAffy() is calling indexProbes() and asking for both PM and MM indices. Since most probesets have no MM probes, you get a bunch of NA values, which is why you see that error.

That's why you are getting errors, but doesn't help you much. So what exactly are you trying to do? But before you answer, please note that the vast majority of the probesets on the miRNA arrays are a set of identical 25-mers that all measure the same 21-23 nt transcript. In other words, the Affy probes are longer than what they measure, so unlike conventional Affy arrays where the probesets are an aggregation of measurements from different positions on the transcript, for the miRNA arrays (for mature miRNAs), all the probes in a probeset are identical. The altcdfenvs package was originally designed to allow people to exclude certain probes from one or more probesets, which might not be relevant in this context.