[BioC] variantAnnotation: alternative GENETIC_CODE, and circular chromosomes?
Valerie Obenchain
vobencha at fhcrc.org
Tue Feb 4 20:38:25 CET 2014
Hi Janet,
Last week we enabled findOverlaps(..., type='within') to work on
circular chromosomes. It was this restriction that prevented
locateVariants() and predictCoding() from handling ChrM.
VariantAnnotation 1.9.34 in devel has the most recent changes.
The output of locateVariants() includes ChrM because the function is
reporting where the range falls wrt the gene (coding, utr, intron,
etc.). predictCoding() however only reports coding variants. If the
annotation you supply does not have any coding regions for ChrM or if
the ranges don't fall in the coding regions then none will be reported.
To confirm your annotation has coding regions for ChrM:
cds <- cdsBy(txdb)
cds[seqnames(cds) %in% "chrM"] ## or whatever the proper name is
#1:
I've added support for the 'genetic.code' and 'if.fuzzy.codon' args to
translate(). To use a different genetic code just pass the named arg
('genetic.code') to predictCoding().
#2:
We've tried to handle the issue of ChrM through making findOverlaps()
behave appropriatly with circular chromosomes. The 'type' argument has
several options (start, end, any, within, equal). This allows quite a
bit of flexibility. When the annotation has an ORF that spans the
start/end findOverlaps() will still behave appropriately according to
'type'.
## annotation with seqlength 9
genes <- GRanges(seqnames=rep.int("A", 4),
IRanges(start=c(2, 4, 6, 8), width=3))
seqinfo(genes) <- Seqinfo(seqnames="A", seqlengths=9, isCircular=TRUE)
## both ranges span the start/end
ranges <- GRanges(seqnames=rep.int("A", 2),
IRanges(9, width=c(2, 4)))
>> findOverlaps(ranges, genes, type="any")
> Hits of length 3
> queryLength: 2
> subjectLength: 4
> queryHits subjectHits
> <integer> <integer>
> 1 1 4
> 2 2 1
> 3 2 4
>> findOverlaps(ranges, genes, type="within")
> Hits of length 1
> queryLength: 2
> subjectLength: 4
> queryHits subjectHits
> <integer> <integer>
> 1 1 4
With the combination of findOverlaps() now working on circular
chromosomes for all values of 'type' and Herve adding the new genetic
codes to Biostrings there should be no need to ignore ChrM. If you run
into trouble or if anything look strange please let us know.
Thanks.
Valerie
On 02/04/2014 04:08 AM, Hervé Pagès wrote:
> Hi Janet,
>
> On 02/03/2014 07:47 PM, Janet Young wrote:
>> Hi there, (I think it'll probably be Valerie looking at this question
>> - hi Valerie),
>>
>> I'm just beginning to look at using VariantAnnotation to annotate some
>> SNPs I've called on some yeast data (sacCer3). I can see this will be
>> a really useful package for me - thanks!
>>
>> I can see that chrM (mitochiondrial) SNPs are currently not included
>> in the output of predictCoding, and then using locateVariants, all of
>> chrM SNPs get annotated as intergenic/NA (with a warning, that we
>> ignore circular chromosomes). I can understand why that is - circular
>> chromosomes, and a different genetic code make it trickier. Fair enough.
>>
>> I'm wondering what the prospects are regarding chrM SNPs in the future
>> - any plans to include those later?
>>
>> I'm also wondering whether I can use some hacks to get chrM SNPs
>> annotated. Two questions/potential issues related to that I wanted to
>> ask you guys about:
>>
>> 1. are alternative codon tables already supported anywhere in
>> Bioconductor? Using "?GENETIC_CODE" it looks like this is defined
>> in Biostrings, and it looks like only the standard nuclear code is
>> defined. Are the various alternative genetic codes defined anywhere?
>> For this project, I'm interested in the yeast mitochondrial code, and
>> for another I'm interested in the fly mitochondrial code. It'd be
>> great if we could have all the codes available (I've got another
>> project looking at ciliate nuclear sequences, for example - not
>> working with translations yet, but maybe later...)
>>
>> With a little work, I'll be able to save flat files from NCBI
>> (http://www.ncbi.nlm.nih.gov/Taxonomy/Utils/wprintgc.cgi), and read
>> those in and transform them to a character vector that looks like
>> GENETIC_CODE. But I realise it might be something useful to have
>> encoded more centrally, so thought I'd ask.
>
> What a timely question! I'll let Val answer the questions about
> support of mitochiondrial DNA in predictCoding() but I can answer
> that particular one. Last week I added a bunch of non standard genetic
> codes to Biostrings (2.31.12). To get the genetic code for Yeast
> Mitochondrial, do:
>
> > getGeneticCode("SGC2")
> TTT TTC TTA TTG TCT TCC TCA TCG TAT TAC TAA TAG TGT TGC TGA TGG CTT
> CTC CTA CTG
> "F" "F" "L" "L" "S" "S" "S" "S" "Y" "Y" "*" "*" "C" "C" "W" "W" "T"
> "T" "T" "T"
> CCT CCC CCA CCG CAT CAC CAA CAG CGT CGC CGA CGG ATT ATC ATA ATG ACT
> ACC ACA ACG
> "P" "P" "P" "P" "H" "H" "Q" "Q" "R" "R" "R" "R" "I" "I" "M" "M" "T"
> "T" "T" "T"
> AAT AAC AAA AAG AGT AGC AGA AGG GTT GTC GTA GTG GCT GCC GCA GCG GAT
> GAC GAA GAG
> "N" "N" "K" "K" "S" "S" "R" "R" "V" "V" "V" "V" "A" "A" "A" "A" "D"
> "D" "E" "E"
> GGT GGC GGA GGG
> "G" "G" "G" "G"
>
> Its format is the same as for GENETIC_CODE. See ?GENETIC_CODE for
> the details.
>
> I also added the 'genetic.code' arg to translate() so you can supply
> an alternate genetic code to use for translation. See ?translate for
> the details.
>
> Please let me know if you find any issues, have questions, or want
> to suggest improvements to these new features.
>
> Thanks,
> H.
>
>>
>> 2. What issues should I think about for the circular chromosomes?
>> I'm thinking of a slightly hacky solution where I ignore any
>> annotated ORFs that wrap around from the end of the chromosome to the
>> beginning, and then just treating it as a linear chromosome.
>> Actually, in my case (using sacCer3) there are no ORFs spanning the
>> break in the circular chromosome, so I don't think I'll miss any
>> annotations. Turns out the same is true for human (hg19 knownGene
>> annotations), so maybe the circular chromosome issue isn't such a big
>> issue after all?
>>
>> It seems like that should work, but any thoughts from you - you've
>> thought about these questions a lot more than I have?
>>
>> Looking forward to hearing any thoughts you have. I know sometimes
>> people just ignore the chrM SNPs, but it'd be nice to take a slightly
>> more comprehensive approach if possible.
>>
>> thanks in advance for any input you have,
>>
>> Janet
>>
>>
>> -------------------------------------------------------------------
>>
>> Dr. Janet Young
>>
>> Malik lab
>> http://research.fhcrc.org/malik/en.html
>>
>> Fred Hutchinson Cancer Research Center
>> 1100 Fairview Avenue N., A2-025,
>> P.O. Box 19024, Seattle, WA 98109-1024, USA.
>>
>> tel: (206) 667 4512
>> email: jayoung ...at... fhcrc.org
>>
>> -------------------------------------------------------------------
>>
>>
>>
>> [[alternative HTML version deleted]]
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at r-project.org
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>> Search the archives:
>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>
>
--
Valerie Obenchain
Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M1-B155
P.O. Box 19024
Seattle, WA 98109-1024
E-mail: vobencha at fhcrc.org
Phone: (206) 667-3158
Fax: (206) 667-1319
More information about the Bioconductor
mailing list