Where to get mouse training set (hapmap/omni) for GATK:VariantRecalibrator

[freshtuo]

I was calling SNP on mouse samples using GATK and was in the step of "Variant quality score recalibration". The VariantRecalibrator walker asked for training sets for mouse SNPs.

Please provide sets of known polymorphic loci marked with the training=true ROD binding tag. For example, -resource:hapmap,known=false,training=true,truth=true,prior=15.0 hapmap_3.3.b37.sites.vcf
-resourcemni,known=false,training=true,truth=false,prior=12.0 1000G_omni2.5.b37.sites.vcf

Does anyone know where to get those training files (specific for UCSC mm9 assembly)? Thanks!

[vivek_]

For mouse you can use the dbSNP. I don't know of any other publicly available resources like Hapmap.

[freshtuo]

Thanks for the quick reply.

I downloaded the dbsnp file (snp128.txt) from http://hgdownload.cse.ucsc.edu/goldenPath/mm9/database/ and converted it into vcf file.

Then I fed that file into the VariantRecalibrator Walker.

java -Xmx4g -jar GenomeAnalysisTK.jar \
-T VariantRecalibrator \
-R Refseq.fa \
-input snps.raw.vcf \
-resource:dbsnp,known=true,training=false,truth=false,prior=6.0 snp128.vcf \
-an QD -an HaplotypeScore -an MQRankSum -an ReadPosRankSum -an FS -an MQ -an InbreedingCoeff \
-mode BOTH \
-recalFile output.recal \
-tranchesFile output.tranches \
-rscriptFile output.plots.R

The program exited with error messages:

##### ERROR MESSAGE: Invalid command line: No training set found! Please provide sets of known polymorphic loci marked with the training=true ROD binding tag. For example, -resource:hapmap,VCF,known=false,training=true,truth=true,prior=12.0 hapmapFile.vcf
##### ERROR ------------------------------------------------------------------------------------------

I was wondering if I can change the parameters by setting both the training/truth to true:

-resource:dbsnp,known=true,training=true,truth=true,prior=6.0 snp128.vcf

Am I doing in the right way? other solutions? Thanks!

Tuo

For mouse you can use the dbSNP. I don't know of any other publicly available resources like Hapmap.

[dpryan]

Depending on exactly what you're doing (i.e. trying to find SNPs within or between strains) the dbSNP may or may not be helpful. For mouse, it lists differences between other strains and the C57Bl/6J reference assembly. If you're interested in SNPs within strains, you'll probably need to bootstrap (i.e. run the process more than once). Also, if you are looking for differences between some strain and the reference, I'm not sure how useful dbSNP would be unless a highly related strain is included (and if so I would filter out all but those SNPs).