This is a test version of Biostars. For the public version, visit https://www.biostars.org.
vcfeval the same variants into both fp.vcf.gz and fn.vcf.gz

Hi

I was trying to do vcfeval to do the concordance between two vcf files. command line I used:

/hgsc_software/java/jdk1.8.0_74/bin/java -jar /hgsc_software/rtg-tools/rtg-tools-3.9.1/RTG.jar vcfeval --threads=1 --baseline=${FILEB}.vcf.gz --calls=${FILEA}.vcf.gz --template=/stornext/snfs5/next-gen/scratch/farek/misc/nist/sdf/GRCh38.SDF --output=${FILEB}-result --ref-overlap --decompose

found out that some matched variants (mostly INDELs) considered not concord and added into both fq.vcf.gz and fn.vcf.gz output files.

Then, out of curiosity, I even concord two exact the same vcf files, suprisingly, I didn't get 100% match, about 2K INDELs added into fp and fn output files, one finding is that these sites' reference concordance overlap with the other closeby called variants. Below is one example:

chr1    243972053   .   CTCTTTCTTCTTTCTTTCTTTCTTTCTT    CTCTT,C 215.02  PASS (this variant got into both fp and fn output files after vcfeval twe exact same vcf)
chr1    243972059   .   C   CT  16.08   PASS (this varinat also called, and consider match and concord)

Even though these two sites had reference overlap, I thouhght by using --ref-overlap, both should concord since I am vcfeval concording the two exact same vcf files.

Really need someone's help and come up with a solution. Thanks

snp

0 answers

No answers yet.

Log in to answer this question.