VarDict Amplicon-Aware Calling with barcoded seq data

Question

[jsneaththompson]

I'm trying out different variant calling software on non-matched somatic samples, where the pcr has been performed using an Agilent kit called Haloplex HS.

At the moment I'm trying VarDict, and I want to try to run VarDict in amplicon-aware mode. The problem is, amplicon-aware mode requires a bed file with start and end positions both with primers and without primers, and Haloplex HS doesn't use standard pcr primers. As can be seen in this image, Haloplex has a molecular barcodes and indexes in addition to a generic pcr primer.

Is it possible to run amplicon-aware calling on barcoded-data? I've tried adding two columns to my amplicons bed file which are the original amplicon start/end points moved one base back/forward, but I get the following error:

Exception in thread "main" htsjdk.samtools.SAMException: Malformed query; start point 139389396 lies after end point 138394717
        at htsjdk.samtools.reference.IndexedFastaSequenceFile.getSubsequenceAt(IndexedFastaSequenceFile.java:215)
        at com.astrazeneca.vardict.ReferenceResource.retriveSubSeq(ReferenceResource.java:31)
        at com.astrazeneca.vardict.ReferenceResource.getREF(ReferenceResource.java:50)
        at com.astrazeneca.vardict.VarDict.ampVardictNotParallel(VarDict.java:5577)
        at com.astrazeneca.vardict.VarDict.start(VarDict.java:69)
        at com.astrazeneca.vardict.Main.run(Main.java:138)
        at com.astrazeneca.vardict.Main.main(Main.java:25)

Does anyone have any ideas on how to approach this?

Answer 1

[Pierre Lindenbaum]

what's the output of

awk -F '\t' '(int($3)<int($2))' input.bed

if there is any ouput, then your bed is malformed (chromEnd>chromStart)

[jsneaththompson]

There's no output when running that command