Bug? (making Pileup file)

airyspace

Dear GATK team

I tried following commands to validate SNP calling (Comparing with SNP Array data)

$ java -jar /usr/local/GenomeAnalysisTK-2.1-5-gf3daab0/GenomeAnalysisTK.jar -T Pileup --showVerbose -R ./human_g1k_v37.fasta -I Sample.final.bam -o Sample.final.pileup
$ java -Xmx2g -jar /usr/local/GenomeAnalysisTK-2.1-5-gf3daab0/GenomeAnalysisTK.jar -R ./human_g1k_v37.fasta -T VariantsToVCF -o Sample.pileup.vcf --variant:SAMPILEUP Sample.final.pileup --dbsnp /home/database/broadinstitute/1.2/b37/dbsnp_132.b37.vcf

then, I got error message like this

"ERROR MESSAGE: the SAM pileup line didn't have the expected number of tokens (expected = 10, saw = 1 on line = 1 14369 C CCCCCCCCCC GGGCC#CIHG"

$ head Sample.final.pileup
1 14369 C CCCCCCCCCC GGGCC#CIHG

Do you have any idea??

Mark_DePristo

These are just the wrong commands for what you want to do. Look at UnifiedGenotyper. The Pileup format is really for debugging, more than anything.

Mark_DePristo

You can use VariantEval for this. Call SNPs with UnifiedGenotyper and then use VariantEval. Docs are all in the Guide section for both tools

airyspace

Dear Mark.

Thank you for your comment, Then, could you recommend for me what to do if I want to compare the accuracy using SNP array data ? I think it would be easy when UnifiedGenotyper calls SNP which is on SNP array.(get dbSNPID which are detected by GATK and check the array data). I want to check all sites on the array.

Sorry, I wish I could write correct English..