Entering edit mode
Hi,
I perviously used SGSeq on our own server w/o any issue; however, we moved our analysis pipeline to a shared facility and I have been unable to more forward using my previously-working pipeline. Here is my R commands, more or less lifted from the docs:
# loading my sample list
load.d <-read.csv("./lympho_b.txt")
data.r<-getBamInfo(load.d,cores=8)
# importing Gencode annotations
txdb<-importTranscripts("/path/toward/gencode/gencode.v31.annotation.gtf")
seqlevelsStyle(txdb) <- "UCSC"
print("Done!!")
txdb<-convertToTxFeatures(txdb)
sgdb<-convertToSGFeatures(txdb)
At this last step, I always get an error and a core dump with the following error:
* Error in `/cvmfs/soft.computecanada.ca/easybuild/software/2017/avx512/Compiler/gcc7.3/r/3.6.0/lib64/R/bin/exec/R': free(): invalid size: 0x000000002bb79c20 *
Is it SGSeq or is it the R install on the shared server? Here is my sessionInfo():
> sessionInfo()
R version 3.6.0 (2019-04-26)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)
Matrix products: default
BLAS/LAPACK: /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/imkl/2018.3.222/compilers_and_libraries_2018.3.222/linux/mkl/lib/intel64_lin/libmkl_gf_lp64.so
locale:
[1] LC_CTYPE=fr_CA.UTF-8 LC_NUMERIC=C
[3] LC_TIME=fr_CA.UTF-8 LC_COLLATE=fr_CA.UTF-8
[5] LC_MONETARY=fr_CA.UTF-8 LC_MESSAGES=fr_CA.UTF-8
[7] LC_PAPER=fr_CA.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=fr_CA.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] stats4 parallel stats graphics grDevices utils datasets
[8] methods base
other attached packages:
[1] SGSeq_1.18.0 SummarizedExperiment_1.14.1
[3] DelayedArray_0.10.0 BiocParallel_1.18.1
[5] matrixStats_0.54.0 Biobase_2.44.0
[7] Rsamtools_2.0.0 Biostrings_2.52.0
[9] XVector_0.24.0 GenomicRanges_1.36.0
[11] GenomeInfoDb_1.20.0 IRanges_2.18.1
[13] S4Vectors_0.22.0 BiocGenerics_0.30.0
loaded via a namespace (and not attached):
[1] Rcpp_1.0.2 compiler_3.6.0 pillar_1.4.2
[4] prettyunits_1.0.2 progress_1.2.2 GenomicFeatures_1.36.4
[7] bitops_1.0-6 tools_3.6.0 zlibbioc_1.30.0
[10] biomaRt_2.40.3 zeallot_0.1.0 digest_0.6.20
[13] bit_1.1-14 RSQLite_2.1.2 memoise_1.1.0
[16] tibble_2.1.3 lattice_0.20-38 pkgconfig_2.0.2
[19] rlang_0.4.0 igraph_1.2.4.1 Matrix_1.2-17
[22] DBI_1.0.0 GenomeInfoDbData_1.2.1 rtracklayer_1.44.2
[25] httr_1.4.1 stringr_1.4.0 hms_0.5.0
[28] vctrs_0.2.0 bit64_0.9-7 grid_3.6.0
[31] R6_2.4.0 AnnotationDbi_1.46.0 XML_3.98-1.20
[34] magrittr_1.5 blob_1.2.0 backports_1.1.4
[37] GenomicAlignments_1.20.1 RUnit_0.4.32 assertthat_0.2.1
[40] stringi_1.4.3 RCurl_1.95-4.12 crayon_1.3.4
Does this happen every time?
I think it's for anyone to tell who is at fault ... seems something more fundamental than SGSeq, though.
Have you tried running this on a different (third) computer? Your personal one, perhaps, or perhaps a linux machine you provision from AWS or something? You've got one situation where it worked, one where it now doesn't -- time to try to break the tie and see what is the what.
Hi,
This happen every time I try to use SGSeq on this particular system; that has not happen to me on my previous work system, my own laptop or my personal server.
My guess is a strange/toxic interaction between SGSeq and this local R install but I need to be sure before I start bugging the sysadmin to try to solve it. I am hoping that the particular error message would give a clue on this failure.
Best regards
S
Perhaps it would be helpful where in the codebase this is triggered?
You might be able to get that info by calling traceback() right after the error is thrown, which hopefully will bring up the function call stack.
Also compare the R and package versions from running sessionInfo() on the failing system and compare that to the machines where it works to ensure you are as close to apples-vs-apples ...