Receiving the following error message anytime I try to run findOverlaps() ...was working before update.
Error in validObject(.Object) :
invalid class “IRanges” object: undefined class for slot "elementMetadata" ("DataTable_OR_NULL")
See Problem with GeneSetCollection after updating R devel packages. and https://stat.ethz.ch/pipermail/bioc-devel/2017-February/010463.html ; you likely need to update some of your installed packages. BiocInstaller::biocValid() might be a start, or the approach in the bioc-devel post.
I assume you are using the 'devel' version of Bioconductor; if not please update your post with the output of sessionInfo(), and add a comment.
Sorry for my super abbreviated post.
I agree that package updating is likely the culprit, but I also can't figure out if the build error on the most recent version of R for GenomicRanges could be causing the problem instead?
I'm currently using R-3.3.1.
Looking for overlaps between two GRanges objects:
GRanges object with 7149 ranges and 6 metadata columns:
GRanges object with 7149 ranges and 0 metadata columns:
> ov <- GenomicRanges::findOverlaps(gtex_regions,sra_regions)
Error in validObject(.Object) :
invalid class “PartitioningByEnd” object: undefined class for slot "elementMetadata" ("DataTable_OR_NULL")
Thanks for the first suggestion. Definitely a good check and something I hadn't done.:
Lastly, I was having the issue in R-devel also ....but I'll try working there. For now, here is sessionInfo() in case something is glaringly obvious...:
R version 3.3.1 Patched (2016-09-30 r71426)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Red Hat Enterprise Linux Server release 6.6 (Santiago)
 LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
 LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
 LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
 LC_PAPER=en_US.UTF-8 LC_NAME=C
 LC_ADDRESS=C LC_TELEPHONE=C
 LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
attached base packages:
 parallel stats4 stats graphics grDevices datasets utils
 methods base
other attached packages:
 BiocInstaller_1.24.0 genefilter_1.56.0
 minfi_1.20.2 bumphunter_1.14.0
 locfit_1.5-9.1 iterators_1.0.8
 foreach_1.4.3 Biostrings_2.42.1
 XVector_0.14.0 SummarizedExperiment_1.4.0
 Biobase_2.34.0 dplyr_0.5.0
 purrr_0.2.2 readr_1.0.0
 tidyr_0.6.1 tibble_1.2
 ggplot2_2.2.1 tidyverse_1.1.1
 GenomicRanges_1.26.2 GenomeInfoDb_1.10.3
 IRanges_2.8.1 S4Vectors_0.12.1
loaded via a namespace (and not attached):
 nlme_3.1-131 bitops_1.0-6 matrixStats_0.51.0
 lubridate_1.6.0 RColorBrewer_1.1-2 httr_1.2.1
 tools_3.3.1 doRNG_1.6 nor1mix_1.2-2
 R6_2.2.0 DBI_0.5-1 lazyeval_0.2.0
 colorspace_1.3-2 mnormt_1.5-5 base64_2.0
 curl_2.3 preprocessCore_1.36.0 rvest_0.3.2
 xml2_1.1.1 pkgmaker_0.22 rtracklayer_1.34.1
 scales_0.4.1 psych_1.6.12 quadprog_1.5-5
 stringr_1.1.0 digest_0.6.12 Rsamtools_1.26.1
 foreign_0.8-67 illuminaio_0.16.0 siggenes_1.48.0
 GEOquery_2.40.0 limma_3.30.10 readxl_0.1.1
 RSQLite_1.1-2 jsonlite_1.2 mclust_5.2.2
 BiocParallel_1.8.1 RCurl_1.95-4.8 magrittr_1.5
 Matrix_1.2-8 Rcpp_0.12.9 munsell_0.4.3
 stringi_1.1.2 MASS_7.3-45 zlibbioc_1.20.0
 plyr_1.8.4 grid_3.3.1 forcats_0.2.0
 lattice_0.20-34 haven_1.0.0 splines_3.3.1
 annotate_1.52.1 multtest_2.30.0 GenomicFeatures_1.26.2
 hms_0.3 beanplot_1.2 rngtools_1.2.4
 reshape2_1.4.2 codetools_0.2-15 biomaRt_2.30.0
 XML_3.98-1.5 data.table_1.10.4 modelr_0.1.0
 openssl_0.9.6 gtable_0.2.0 reshape_0.8.6
 assertthat_0.1 xtable_1.8-2 broom_0.4.1
 survival_2.40-1 GenomicAlignments_1.10.0 AnnotationDbi_1.36.2
 registry_0.3 memoise_1.0.0
The even numbered 'y' of x.y.z versions in your Bioconductor packages indicate that you are working in the 'release' branch, so the build failure in the devel branch (the one that you pointed to, Bioconductor version 3.5) is not relevant. Also, the error that you reported, about DataTable_OR_NULL would only come from a devel package. So I guess the sessionInfo() you report above is not from the session where the bug was observed, or you've sometimes installed devel packages and then replaced them with release packages, or...?
Using biocLite() and the procedures recommended on each package landing page is a good way to avoid problems. Following the guidelines on http://bioconductor.org/developers/how-to/useDevel/ (these guidelines depend on release cycle) is the right way to use the devel version of Bioconductor.
Login before adding your answer.
Use of this site constitutes acceptance of our User Agreement and Privacy