Why this SNP was classified as FN by vcfeval?

[L Mao]

Hi,

I used the following command to compare my VCF (file "onco1.snp.vcf.gz" which was called from an NA12878 sample) with NA12878 standard (file 'benchmark.snp.hg19.vcf.gz'). File "oncorisk.isHC.bed" is an intersect of target region bed file and NA12878 benchmark bed file.

rtg vcfeval -b benchmark.snp.hg19.vcf.gz -c onco1.snp.vcf.gz -e oncorisk.isHC.bed -t ref/sdt --all-records -o outdir

A FN (False Negative) SNP was found in an output file "fn.vcf.gz":

chr7    142460313    rs6666    T    C    50    PASS    platforms=4;platformnames=PacBio,Illumina,10X,Ion;datasets=4;datasetnames=CCS15kb_20kb,HiSeqPE300x,10XChromiumLR,IonExome;callsets=6;callsetnames=CCS15kb_20kbDV,CCS15kb_20kbGATK4,HiSeqPE300xGATK,10XLRGATK,HiSeqPE300xfreebayes,IonExomeTVC;datasetsmissingcall=CGnormal,SolidSE75bp;callable=CS_CCS15kb_20kbDV_callable,CS_CCS15kb_20kbGATK4_callable;filt=CS_CGnormal_filt;difficultregion=HG001.hg37.300x.bam.bilkentuniv.010920.dups,hg19.segdups_sorted_merged,mm-2-merged    GT:PS:DP:ADALL:AD:GQ  1/1:.:846:0,249:1,79:224

However, this SNP was included in my input vcf file "onco1.snp.vcf.gz":

chr7    142460313    rs6666    T    C    66571.64    .    AC=1;AF=0.500;AN=2;BaseQRankSum=-10.035;DB;DP=6468;ExcessHet=3.0103;FS=4.067;MLEAC=1;MLEAF=0.500;MQ=55.56;MQRankSum=42.836;QD=10.51;ReadPosRankSum=-7.877;SOR=1.022    GT:AD:DP:GQ:PL    0/1:3424,2913:6337:99:66579,0,127995

The depth and quality at this position was high as shown in the vcf line above.

So why RTG Tools classified it as a FN? Thanks.

[Sean Irvine]

Hi,

The baseline file (benchmark.snp.hg19.vcf.gz) defines this locus to be homozygous with GT 1/1, but in your sample file (onco1.snp.vcf.gz) the locus is heterozygous with GT 0/1.  Therefore, the calls file only provides evidence for one C at this locus, so the records are not a strict match.

If you don't care about the zygosity of the matches, then you can give vcfeval the --squash-ploidy option.  There is more detail about this in the user manual.

Regards,

Sean.

--
You received this message because you are subscribed to the Google Groups "RTG Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rtg-users+...@realtimegenomics.com.
To view this discussion on the web visit https://groups.google.com/a/realtimegenomics.com/d/msgid/rtg-users/08a18731-4762-44c3-a6b0-20acb6742354n%40realtimegenomics.com.

[L Mao]

I see. Thanks Sean!