Error in creating dir

[H-Qiya]

Hi, Mr Mertes and your team! I've read your manual and relative literature and found your package really worth a try of best effort! However, when running your example data as well as my own data, I met the same error from the very first step as follows: #######Example data in your manual#######

settings<-FraserDataSet(colData=sampleTable,workingDir=tempdir()) settings -------------------- Sample data table -----------------

A tibble: 3 x 5

sampleID bamFile group gene pairedEnd

1 sample1 extdata/bam/sample1.bam 1 TIMMDC1 TRUE 2 sample2 extdata/bam/sample2.bam 3 CLPP TRUE 3 sample3 extdata/bam/sample3.bam 2 MCOLN1 TRUE

----------------------- Settings ----------------------- Analysis name: Data Analysis Analysis is strand specific: no Working directory: 'C:\Users\ADMINI~1\AppData\Local\Temp\RtmpQ9cxvH'

-------------------- BAM parameters -------------------- Fri Dec 11 09:26:15 2020: The given working directory 'C:\Users\biocbuild\bbs-3.12-bioc\tmpdir\Rtmpw1fLb2/FRASER' does not exists. We will create it. Error in h(simpleError(msg, call)) : error in evaluating the argument 'object' in selecting a method for function 'scanBamParam': invalid class “FraserDataSet” object: Make sure we can write to the given working directory ' C:\Users\biocbuild\bbs-3.12-bioc\tmpdir\Rtmpw1fLb2/FRASER '. In addition: Warning message: In dir.create(object@workingDir, recursive = TRUE) : cannot create dir 'C:\Users\biocbuild', reason 'Permission denied'

When working on my own data, I set my own working directory at the first step, and other script is almost the same. Nonetheless, the same error still exists: #######My own data####### ----------------------- Settings ----------------------- Analysis name: Data Analysis Analysis is strand specific: no Working directory: 'Y:/HCM-newly-sequenced/200820-WGX' -------------------- BAM parameters -------------------- Fri Dec 11 09:51:35 2020: The given working directory 'C:\Users\biocbuild\bbs-3.12-bioc\tmpdir\Rtmpw1fLb2/FRASER' does not exists. We will create it. Error in h(simpleError(msg, call)) : error in evaluating the argument 'object' in selecting a method for function 'scanBamParam': invalid class “FraserDataSet” object: Make sure we can write to the given working directory ' C:\Users\biocbuild\bbs-3.12-bioc\tmpdir\Rtmpw1fLb2/FRASER '. In addition: Warning message: In dir.create(object@workingDir, recursive = TRUE) : cannot create dir 'C:\Users\biocbuild', reason 'Permission denied'

Is the working directory completely fixed in ' C:\Users\biocbuild\bbs-3.12-bioc\tmpdir\Rtmpw1fLb2/FRASER ' ? How can I alter it into my own path? Otherwise, must I run R as an administrator? Yet this might lead to unpredictable risk? Looking forward to your reply! Thanks a lot!

[c-mertes]

Dear @H-Qiya, unfortunately, I can not reproduce your error. Even on Windows, I can not reproduce it. Which version do you use and how did you install it? Could you please try the code below in a fresh session?

You do not need to be an administrator and you set already correctly the working dir. So I'm very curious which version you use.

BiocManager::install("FRASER")

library(FRASER)

sampleTable <- fread(system.file("extdata", "sampleTable.tsv", package = "FRASER", mustWork = TRUE))
bamFiles <- system.file(sampleTable[, bamFile], package = "FRASER", mustWork = TRUE)
sampleTable[, `:=`(bamFile, bamFiles)]

settings<-FraserDataSet(colData=sampleTable,workingDir=tempdir())
settings

BiocManager::valid()

This is run under R 4.0.2 and Bioconductor version 3.11 and FRASER 1.0.2

[H-Qiya]

Hi, @c-mertes ! Thanks a lot for your reply! What is unbelievable is that the Error in "---Bam parameters---" disappeared after copying your code above and running in R studio to reinstall FRASER! Meanwhile, my version is R 4.0.3, Bioconductor 3.12 and FRASER 1.2.0. And the code which reported error last time was copied from FRASER manual pdf. Last time, I ran example data and my data respectively in R as an administrator to avoid the above "---Bam parameters---" error, though it was risky. And this time I ran the same data using FRASER without administrator mode. Then I compared those results and found they are exactly the same (I kept set.seed parameters consistent). Anyway, this error has been fixed, though I didn’t figure out how it is produced deeply in machine, hhh.

Also, I noticed your article has been published on NC the other day(DOI: 10.1038/s41467-020-20573-7). Congratulations on your achievement!!!

[c-mertes]

Thanks for the feedback and for the congratulations! As it is resolved now, I will close the bug. Please let me know if anything else is going wrong in the future!