Merging gvcf files between Dragen and other calling tools

Question

[tacrolimus]

Dear all,

Does anyone have experience merging DRAGEN called gvcfs and those called by other softwares. DRAGEN adds the "NON_REF" tag to many of it's variants INFO columns which then makes merging and annotation with non-DRAGEN called gvcfs tricky e.g.:

GL000192.1    546636    .    G    A,<NON_REF>    49.77    .    BaseQRankSum=1.026;ClippingRankSum=1.026;DP=4;MLEAC=1,0;MLEAF=0.500,0.00;MQ=49.02;MQ0=0;MQRankSum=-1.026;ReadPosRankSum=0.000    GT:AD:DP:GQ:PGT:PID:PL:SB    0/1:2,2,0:4:78:0|1:546636_G_A:78,0,87,84,93,177:1,1,1,1
GL000192.1    546637    .    G    <NON_REF>    .    .    END=546645    GT:DP:GQ:MIN_DP:PL    0/0:4:11:4:0,12,99

I need to merge around 1000 gvcfs (Across a single gene) and would appreciate any insights into this! I am using bcftools via a linux HPC.

All the best

[Pierre Lindenbaum]

Hi tried to use GATK CombineGVCFs a while ago, without success, the INFO are not compatible.

[tacrolimus]

Did you not merge them all then or just rid of the INFO entirely? Many thanks!

[Pierre Lindenbaum]

those are GVCF files, not standard VCF files, you cannot merge those files without using the INFO column.

[tacrolimus]

Dear Pierre,

If one set of gvcf files are ISAAC aligned (>80,000 samples) and the other DRAGEN (100 samples) as discussed above but I have access to the original CRAM files for the DRAGEN files in theory could I recall the DRAGEN files into gVCF files that are able to be merged with the other files?

Many thanks