[BioC] curious behavior from VariantAnnotation 1.5.36

[Valerie Obenchain]

Hi Tim,

Thanks for reporting this. There was a typo in the method called for 
3UTR's in the AllVariants() code. Now fixed in 1.5.38 and 1.4.9.

library(Homo.sapiens)
library(VariantAnnotation)
library(TxDb.Hsapiens.UCSC.hg19.knownGene)
txdb <- TxDb.Hsapiens.UCSC.hg19.knownGene
site <- GRanges(c('chr1'), IRanges(start=36521401, width=1))
EIF2C3 <- unlist(mget('EIF2C3', org.Hs.egSYMBOL2EG))

all <- locateVariants(site, txdb, AllVariants())
three <- locateVariants(site, txdb, ThreeUTRVariants())

 > identical(all, three)
[1] TRUE


Valerie




On 02/13/2013 10:19 PM, Tim Triche, Jr. wrote:
> Somehow I omitted a line... this should go immediately after
> library(VariantAnnotation), i.e.
>
>
> library(Homo.sapiens)
> library(VariantAnnotation)
>
> ## this following line is the one that went missing
> EIF2C3 <- unlist(mget('EIF2C3', org.Hs.egSYMBOL2EG))
>
> EIF2C3.site <- GRanges(c('chr1'), IRanges(start=36521401, width=1))
> EIF2C3.site %over%
> transcriptsBy(TxDb.Hsapiens.UCSC.hg19.knownGene)[[EIF2C3]]
>
> ## ...the rest is as before
>
>
> On Wed, Feb 13, 2013 at 5:09 PM, Tim Triche, Jr. <tim.triche at gmail.com>wrote:
>
>> I was reannotating some discriminating loci today and found this odd case
>> -- a site is in the 3' UTR of two transcripts for a gene, and
>> threeUTRVariants() finds it, but AllVariants() does not.
>>
>> Code to reproduce the result at an example locus:
>>
>> packageVersion('VariantAnnotation')
>> ## [1] '1.5.36'
>> packageVersion('Homo.sapiens')
>> ## [1] '1.0.0'
>>
>> ## test case for bioc-list:
>> library(Homo.sapiens)
>> library(VariantAnnotation)
>> EIF2C3.site <- GRanges(c('chr1'), IRanges(start=36521401, width=1))
>> EIF2C3.site %over%
>> transcriptsBy(TxDb.Hsapiens.UCSC.hg19.knownGene)[[EIF2C3]]
>> ## TRUE
>>
>> allVars <- locateVariants(EIF2C3.site,
>>                            TxDb.Hsapiens.UCSC.hg19.knownGene,
>>                            AllVariants())
>> length(allVars)
>> ## 0
>>
>> threePrimeVars <- locateVariants(EIF2C3.site,
>>                                   TxDb.Hsapiens.UCSC.hg19.knownGene,
>>                                   ThreeUTRVariants())
>> length(threePrimeVars)
>> ## 2
>>
>> Why aren't the 3' UTR matches included in the AllVariants() results?
>> That doesn't seem like the expected behavior, at least not to me.
>>
>>
>>
>> R> sessionInfo()
>> R Under development (unstable) (2013-02-13 r61937)
>> Platform: x86_64-unknown-linux-gnu (64-bit)
>>
>> locale:
>>   [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
>>   [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
>>    [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
>>   [7] LC_PAPER=C                 LC_NAME=C
>>   [9] LC_ADDRESS=C               LC_TELEPHONE=C
>> [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
>>
>> attached base packages:
>> [1] parallel  stats     graphics  grDevices datasets  utils     methods
>> [8] base
>>
>> other attached packages:
>>   [1] FDb.InfiniumMethylation.hg19_1.0.6
>>   [2] BSgenome.Hsapiens.UCSC.hg19_1.3.19
>>   [3] BSgenome_1.27.1
>>   [4] Homo.sapiens_1.0.0
>>   [5] TxDb.Hsapiens.UCSC.hg19.knownGene_2.8.0
>>    [6] org.Hs.eg.db_2.8.0
>>   [7] GO.db_2.8.0
>>   [8] RSQLite_0.11.2
>>   [9] DBI_0.2-5
>> [10] OrganismDbi_1.1.13
>> [11] VariantAnnotation_1.5.36
>> [12] Rsamtools_1.11.16
>> [13] Biostrings_2.27.11
>> [14] GenomicFeatures_1.11.8
>>   [15] AnnotationDbi_1.21.10
>> [16] Biobase_2.19.2
>> [17] chromophobe_0.50
>> [18] pheatmap_0.7.4
>> [19] ggplot2_0.9.3
>> [20] reshape2_1.2.2
>> [21] GenomicRanges_1.11.29
>> [22] IRanges_1.17.31
>> [23] BiocGenerics_0.5.6
>>   [24] BiocInstaller_1.9.6
>> [25] gtools_2.7.0
>> [26] devtools_1.1
>>
>> loaded via a namespace (and not attached):
>>   [1] biomaRt_2.15.0     bitops_1.0-5       colorspace_1.2-1
>> dichromat_2.0-0
>>   [5] digest_0.6.2       evaluate_0.4.3     graph_1.37.5       grid_3.0.0
>>
>>   [9] gtable_0.1.2       httr_0.2           labeling_0.1
>> lattice_0.20-13
>> [13] MASS_7.3-23        Matrix_1.0-10      memoise_0.1        munsell_0.4
>>
>> [17] plyr_1.8           proto_0.3-10       RBGL_1.35.0
>>   RColorBrewer_1.0-5
>> [21] RCurl_1.95-3       rtracklayer_1.19.9 scales_0.2.3       stats4_3.0.0
>>
>> [25] stringr_0.6.2      tcltk_3.0.0        tools_3.0.0        whisker_0.1
>>
>> [29] XML_3.95-0.1       zlibbioc_1.5.0
>>
>>
>> --
>> *A model is a lie that helps you see the truth.*
>> *
>> *
>> Howard Skipper<http://cancerres.aacrjournals.org/content/31/9/1173.full.pdf>
>>
>
>
>