Using R-devel for windows, I attempted to run the posted example script for the PoTRa R program, and one of its dependencies is the graphite bioconductor package (graphite is linked to the error).  PoTRa's functions, example data and example code, plus some detail can be found here:

http://dinulab.org/tools/potra/

So my main request is if anyone can help me troubleshoot this error message. The developers of PoTRa also do not understand why the error messages below appear on some computers but not in others.

The error messages are the following:

> ##For PoTRA.cor:
>
> results.cor <-PoTRA.cor(mydata=mydata,genelist=genelist,Num.sample.normal=8,Num.sample.case=8,Pathway.database=humanKEGG[1:10],PR.quantile=0.95)
[1] 1
Error in nrow(edges) :
  no slot of name "protEdges" for this object of class "Pathway"

> ##For PoTRA.comb:
>
> results.comb <-PoTRA.comb(mydata=mydata,genelist=genelist,Num.sample.normal=8,Num.sample.case=8,Pathway.database=humanKEGG[1:10],PR.quantile=0.95)
[1] 1
Error in nrow(edges) :
  no slot of name "protEdges" for this object of class "Pathway"
 

Here is a link to the graphite info page where protEdges is a slot for Pathway. Note that the 3.6 version issued the same error messages as the 3.7 version in bioconductor:

https://rdrr.io/bioc/graphite/src/R/pathway.R

Code block from pathway.R from Graphite

setClass("Pathway",
  representation(
    id = "character",
    title = "character",
    database = "character",
    species = "character",
    protEdges = "data.frame",
    protPropEdges = "data.frame",
    metabolEdges = "data.frame",
    metabolPropEdges = "data.frame",
    mixedEdges = "data.frame",
    timestamp = "Date"))

Here is my sessionInfo():

> sessionInfo()
R Under development (unstable) (2018-03-10 r74380)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 16299)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252    LC_MONETARY=English_United States.1252 LC_NUMERIC=C                           LC_TIME=English_United States.1252    

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

other attached packages:
[1] org.Hs.eg.db_3.5.0   AnnotationDbi_1.41.4 IRanges_2.13.28      S4Vectors_0.17.35    Biobase_2.39.2       igraph_1.2.1         graphite_1.25.1      graph_1.57.1         BiocGenerics_0.25.3

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.15    magrittr_1.5    rappdirs_0.3.1  bit_1.1-12      R6_2.2.2        rlang_0.2.0     blob_1.1.0      httr_1.3.1      checkmate_1.8.5 DBI_0.8         bit64_0.9-7     digest_0.6.15  
[13] tibble_1.4.2    memoise_1.1.0   RSQLite_2.0     compiler_3.5.0  pillar_1.2.1    backports_1.1.2 pkgconfig_2.0.1

Dear user,

I'm one of the authors of the graphite package. The error you report depends on the fact that the example rdata provided on the PoTRA homepage erroneously includes an old copy of graphite pathways (object humanKEGG), instead of retrieving them from the package itself. The internal structure of the Pathway class we define has changed lately and R thus becomes confused as it finds on your system the latest version of the package, but an older version of the data. Hence the mismatch in the slots. The proper solution is to stop using the duplicated pathway data or, even better, to organize PoTRA as a proper package so that it can explicitly depends on graphite.

Best regards,

  Gabriele Sales