[BioC] GWASTools

[Stephanie M. Gogarten]

Dear Uli,

Thanks for the suggestion.  In GWASTools 1.5.9 (which will become the 
next release very soon), the "snp.annot" argument in convertNcdfGds 
should be a SnpAnnotationDataFrame, and the chromosome codes from 
snp.annot will be preserved in the resulting GDS file.

best wishes,
Stephanie

On 1/16/13 2:42 AM, Ulrich Knief wrote:
> Dear Stephanie,
> I entered this code and it seems it changes the internal chromosome
> coding in SNPRelate after using 'convertNcdfGds'.
>
> path = ""
> geno.nc.file <- "tmp.geno.nc"
> ncfile <- paste(path,geno.nc.file,sep="")
> gdsfile <- "tmp_geno.gds"
> annot1 <- data.frame(pData(scanAnnot1))
> convertNcdfGds(ncfile, gdsfile, sample.annot=annot1)
>
> ZFwild <- openfn.gds("tmp_geno.gds", readonly=FALSE)
>
> # change internal chromosome coding
>     option <- snpgdsOption(autosome.start=1, autosome.end=32)
>     chrcode <- option$chromosome.code
>     chr <- snpAnnot$chromosome
>     for (i in names(chrcode)) chr[snpAnnot$chromosome == i] <- chrcode[[i]]
>     chr <- as.integer(chr)
>     chr[is.na(chr)] <- 0
>     put.attr.gdsn(index.gdsn(ZFwild, c("snp.chromosome")),
> "autosome.start", option$autosome.start)
>     put.attr.gdsn(index.gdsn(ZFwild, c("snp.chromosome")),
> "autosome.end", option$autosome.end)
>     for (i in 1:length(option$chromosome.code)) {
>          put.attr.gdsn(index.gdsn(ZFwild, c("snp.chromosome")),
> names(option$chromosome.code)[i],
>          option$chromosome.code[[i]])
>     }
>     sync.gds(ZFwild)
>
>
> Best wishes
> Uli
>
>
>
> -------- Original Message --------
> Subject: 	Re: GWASTools
> Date: 	Wed, 16 Jan 2013 10:12:16 +0100
> From: 	Ulrich Knief <uknief at orn.mpg.de>
> To: 	Stephanie M. Gogarten <sdmorris at u.washington.edu>
>
>
>
> Dear Stephanie,
> sorry to bother you again. I have a problem with defining new chromosome
> codes during conversion from GWASTools to SNPRelate. I use the function
> 'convertNcdfGds' from the GWASTools package. You cannot specify
> differing chromosome names in this function.
> Then I tried to change the internal chromosome representations in
> SNPRelate with 'snpgdsOption'. But the list output of this function is
> only useable for SNPRelate functions while importing data from PLINK or
> VCF files (snpgdsBED2GDS, snpgdsVCF2GDS) and not for the
> 'convertNcdfGds' function. Am I missing something?
>
> I am not sure, but I think in your tutorial on data cleaning you made a
> mistake on page 47 when plotting heterozygosity in females on the X
> chromosome. Shouldn't it be:
> female <- scanAnnot$sex == "F"
> boxplot(scanAnnot$het.X[female] ~ scanAnnot$population[female], main="X
> Chromosome Heterozygosity in Females")
>
>
> Best wishes
> Uli
>
>
>
>
>
>
> On 1/7/2013 5:46 PM, Stephanie M. Gogarten wrote:
>> "anomDetectLOH" may be what you want.  Be warned that it takes a long
>> time to run.
>>
>> Stephanie
>>
>> On 1/6/13 4:19 PM, Uli wrote:
>>> Dear Stephanie,
>>> thank you very much for your information and support! I really
>>> appreciate your help. I will start writing my scripts in GWASTools and
>>> SNPRelate the next days. If I come across any bugs then I could contact
>>> you again. Are you planning to implement something like PLINKs 'Runs of
>>> Homozygosity' tool in the future?
>>>
>>> Best wishes
>>> Uli
>>>
>>>
>>>
>>> Am 03.01.2013 22:49, schrieb Stephanie M. Gogarten:
>>>> I think Xiuwen may still be on vacation, but I just noticed that the
>>>> latest SNPRelate (0.9.8) has the option to change autosome coding.
>>>> Look at "?snpgdsOption".
>>>>
>>>> Stephanie
>>>>
>>>> On 12/21/12 10:39 AM, Stephanie M. Gogarten wrote:
>>>>> Dear Uli,
>>>>>
>>>>> Xiuwen Zheng is the author of SNPRelate, and as far as I know it still
>>>>> has autosomes hard-coded as 1:22.
>>>>>
>>>>> Xiuwen, do you have any plans to revise SNPRelate so that it can be
>>>>> used
>>>>> with a different number of autosomes?  I just did this in
>>>>> GWASTools, so
>>>>> I can give you an example.
>>>>>
>>>>> Stephanie
>>>>>
>>>>> On 12/20/12 4:23 PM, Uli wrote:
>>>>>> Dear Stephanie,
>>>>>> thank you very much for this update - which was very fast - and for
>>>>>> letting me know! This gives GWASTools a clear advantage over PLINK in
>>>>>> model organisms for all the quality checking. Actually, I think I
>>>>>> will
>>>>>> put together a pipeline that uses GWASTools for quality control
>>>>>> and for
>>>>>> the detection of trisomies, then SNPRelate for PCA and then either
>>>>>> the
>>>>>> emma or the GAPIT package (which includes emma) for the association
>>>>>> testing. Since SNPRelate was written by your working group as well,
>>>>>> does
>>>>>> it also accept the new chromosome names?
>>>>>>
>>>>>> Regards
>>>>>> Uli
>>>>>>
>>>>>>
>>>>>> Am 19.12.2012 21:44, schrieb Stephanie M. Gogarten:
>>>>>>> Dear Dr. Knief,
>>>>>>>
>>>>>>> GWASTools 1.5.5 (available here:
>>>>>>>http://bioconductor.org/packages/2.12/bioc/html/GWASTools.html) now
>>>>>>> allows you to change the chromosome annotations for autosomes.
>>>>>>> There
>>>>>>> is an example on the GenotypeData man page ("?GenotypeData") for how
>>>>>>> to set up a GenotypeData object for a bird species.
>>>>>>>
>>>>>>> Stephanie
>>>>>>>
>>>>>>> On 11/15/12 1:50 AM, Ulrich Knief wrote:
>>>>>>>> Hello Dr Gogarten,
>>>>>>>> I would like to use your R package GWASTools but I am wondering
>>>>>>>> whether
>>>>>>>> it is possible to change the internal chromosome names.
>>>>>>>> Specifically,
>>>>>>>> the organism I am studying has more chromosomes than humans and it
>>>>>>>> would
>>>>>>>> thus be nice to change the chromosome annotations. Since I am
>>>>>>>> working on
>>>>>>>> a bird species it would also be nice to change X to Z but this
>>>>>>>> is not
>>>>>>>> absolutely necessary.
>>>>>>>> Is that possible?
>>>>>>>>
>>>>>>>> Regards,
>>>>>>>> Ulrich Knief
>>>>>>
>>>
>
>
>
> --
> J. Ulrich Knief
> Max Planck Institute for Ornithology
> Department of Behavioural Ecology & Evolutionary Genetics
> Eberhard-Gwinner-Straße, House 7
> 82319 Seewiesen (Starnberg), Germany
> Tel.: +49 (0) 8157 932 - 303
>
>
>
>