Attention:
The frontline support team will be unavailable to answer questions on April 15th and 17th 2019. We will be back soon after. Thank you for your patience and we apologize for any inconvenience!

SelectVariants quick question about concordance/discordance

myoglumyoglu Member
edited November 2013 in Ask the GATK team

Hi!
This is my code:

java -Xmx4g -jar GenomeAnalysisTK-2.7-2-g6bda569/GenomeAnalysisTK.jar \
    -R genome.fa \
    --filter_reads_with_N_cigar \
    -T CombineVariants \
    -V:NORMALSpost RNA-edit_SNPs_NORMALS-post_SNPeff_VA.vfc \
    -V:OMNI 1000G_omni2.5.hg19.vcf \
    -V:db137 dbsnp_137.hg19.vcf \
    -V:Hapmap hapmap_3.3.hg19.vcf \
    -V:ESP1,VCF ESP6500SI-V2-SSA137.updatedRsIds.Allchr.snps_indels_FIX.vcf \
    --out NORMALS-post_TruePositives_raw.vcf &&

java -Xmx4g -jar GenomeAnalysisTK.jar \
    -R genome.fa \
    --filter_reads_with_N_cigar \
    -T SelectVariants \
    -V NORMALS-post_TruePositives_raw.vcf \
    -select "set == 'NORMALSpost'" \
    --out NORMALS-post_TruePositives.vcf

I hope I got this right: I now have SNPs (I have only called SNPs, not indels) that is found in my sample, 1000G, dbSNP and ESP. Right?

Question:
Can I easily make a VCF with all SNPs in MY sample only, not found in anything else by tweaking this code? Or must I do it some way else?

Thanks!

Post edited by Geraldine_VdAuwera on

Best Answer

Answers

  • myoglumyoglu Member

    If there is not I have an alternative strategy:

    Using CombineVariants to "merge" all my references (1000g etc)

    Use SelectVariants discordance option against this "merged" vcf

  • myoglumyoglu Member

    Yes, thats what I finally did. Thanks!

  • n_mahfouzn_mahfouz GermanyMember

    Hi Geraldine,
    I would like to get variants unique to a data set as well, but i was wondering if i could compare against reads rather than other variants ... is there an alternative that applies to my problem??
    Thanks!

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    @n_mahfouz If I'm not mistaken, you also posted this same question as a new thread, right? Please don't do that; we read and respond to everything, so double postings make us waste time.

Sign In or Register to comment.