-L option for <DUP> structural variants

[NEOGrok]

Hi,

We're trying to detect an FLT3 ITD using Vardict, and it does detect it, but I'm not sure why it reports the ALT as \<DUP> given that the ITD length is reported as 285 and the default for the -L option is 1000. I've tried setting it to different values but it doesn't change the output. Perhaps I'm doing something wrong? I've read in a different issue that some aspects of duplicate detection aren't supported yet. Does it include this?

The command I'm running is

vardict-java \
-G $REF_GEN \
-N Horizon \
-b Horizon.bam \
-r 1 \
-R 13:28033800-28033950 \
| teststrandbias.R \
| var2vcf_valid.pl -N tumor -E -f 0.01 \
| awk '{if ($1 ~ /^#/) print; else if ($4 != $5) print}' > Horizon-CMP001.vcf

and the called variant from the VCF file is

13      28033909        .       C       <DUP>   204     PASS    SAMPLE=tumor;TYPE=DUP;DP=884;VD=61;AF=0.069;BIAS=2:2;REFBIAS=485:279;VARBIAS=32:29;PMEAN=30.5;PSTD=1;QUAL=34.5;QSTD=1;SBF=0.09889;ODDRATIO=1.57441;MQ=60;SN=122;HIAF=0.0833;ADJAF=0.069;SHIFT3=0;MSI=0;MSILEN=0;NM=0.1;HICNT=61;HICOV=732;LSEQ=TTCAGCATTTTGACGGCAAC;RSEQ=TGGCTTTCATACCTAAATTG;DUPRATE=0;SVTYPE=DUP;SVLEN=285;SPLITREAD=32;SPANPAIR=29 GT:DP:VD:AD:AF:RD:ALD   0/1:884:61:764,61:0.069:485,279:32,29

Thank you, Francisc

[PolinaBevad]

Hello Francisc, For DUPs we also have another strategy to find them with flanking sequences length, so this option will not change the situation (the variant will be recognized as DUP anyway). It is the only one exclusion, INV/DEL will always use the -L threshold.

[Montana]

This will be using the -L threshold.