getBMFeatureAnnos error

[wikiselev]

Hi Davis,

I am running the getBMFeatureAnnos like this:

sceset <- getBMFeatureAnnos(
    sceset, filters="ensembl_gene_id",
    biomart="ensembl", dataset="mmusculus_gene_ensembl")

and get the following error:

Entity 'copy' not defined
Error: 1: Entity 'copy' not defined

It looks a biomaRt connection problem according to this question, but I am not sure... It started occurring since yesterday. Could you test it?

Vlad

My sessionInfo:

R Under development (unstable) (2016-10-20 r71540)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X Yosemite 10.10.5

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] parallel  stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] scater_1.3.42       ggplot2_2.2.1       Biobase_2.35.1      BiocGenerics_0.21.3

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.10         vipor_0.4.5          plyr_1.8.4          
 [4] zlibbioc_1.21.0      viridis_0.4.0        bitops_1.0-6        
 [7] tools_3.4.0          biomaRt_2.31.5       digest_0.6.12       
[10] rhdf5_2.19.2         RSQLite_1.1-2        memoise_1.0.0       
[13] tibble_1.3.0         gtable_0.2.0         viridisLite_0.2.0   
[16] lattice_0.20-34      Matrix_1.2-8         shiny_1.0.1         
[19] DBI_0.6-1            beeswarm_0.2.3       gridExtra_2.2.1     
[22] stringr_1.2.0        dplyr_0.5.0          knitr_1.15.1        
[25] S4Vectors_0.13.15    IRanges_2.9.19       locfit_1.5-9.1      
[28] stats4_3.4.0         grid_3.4.0           shinydashboard_0.5.3
[31] data.table_1.10.4    R6_2.2.0             AnnotationDbi_1.37.4
[34] XML_3.98-1.6         ggbeeswarm_0.5.3     limma_3.31.19       
[37] reshape2_1.4.2       edgeR_3.17.5         magrittr_1.5        
[40] matrixStats_0.52.1   scales_0.4.1         htmltools_0.3.5     
[43] tximport_1.3.13      assertthat_0.1       mime_0.5            
[46] xtable_1.8-2         colorspace_1.3-2     httpuv_1.3.3        
[49] stringi_1.1.3        RCurl_1.95-4.8       lazyeval_0.2.0      
[52] munsell_0.4.3        rjson_0.2.15        

[LTLA]

Yes, it's a BiomaRt problem. I'm guessing that Ensembl changed something internally, and now some of the expected fields aren't present any more. Might be worthwhile posting a MWE to BioC-support.

Edit: Speaking of which: https://support.bioconductor.org/p/94725/#94729

[wikiselev]

Thanks, Aaron! From the bioconductor forum it's still not clear whether it is just a temporary problem or not and whether you need to modify the getBMFeatureAnnos to adapt to the change.

I have a couple of automatic pipelines that rely on this function (one of them is our scRNA-Seq course), they fail now...

[wikiselev]

Ok, here is some clarification (https://support.bioconductor.org/p/94725/#94781):

Dear all,

I am afraid that we are experiencing some issues with our marts since yesterday's maintenance. We are aiming to fix the issue as soon as possible.

Apologies for any inconvenience caused, Kind Regards, Thomas

[wikiselev]

OK, it works now.

[davismcc]

Ah, yes, the old BiomaRt dependency problem...

I think it would be worthwhile to add an annotation function that would work from an EnsDB object or similar.

Aaron/Vlad: any thoughts on how you would envision such a function to work? What's your preferred approach to annotation (not relying on a network connection to Biomart or other web resources)?

[LTLA]

org.Mm.eg.db and friends.

[wikiselev]

But in this case you will need to install EnsDBs? I've always used biomaRt though and never had problems with it. But if you think EnsDBs would be better I am fine with it.

[davismcc]

I'm not going to scrap the getBMFeatureAnnos() function. For the time being it is fine, but I seem to recall that at the Bioc Europe devel meeting in December someone mentioned that Biomart will be closing down early next year (or something like that). So in the long term users/we/I should have a replacement.

Of course, this is not "core" functionality. Users can pretty easily get annotations from org.Mm.eg.db etc and add them, so I would like a convenience function that takes an SCESet object and an org object and annotates in one step. Missed the boat for this release, though, so i'll close this issue and open up another one as a longer term reminder that I want to add this.