call a vaiant without chr and position, just the following info

[asmlgkj]

AGAAGCCTCTACAAGTGAG GT:DP:VD:ALD:RD:AD:AF:BIAS:PMEAN:PSTD:QUAL:QSTD:SBF:ODDRATIO:MQ:SN:HIAF:ADJAF:NM 0/0:332:2:1,1:161,169:330,2:0.006:2,2:17:0:38:0:1:1.05:60:4:0.0061:0:1 0/0:283:3:3,0:133,145:278,3:0.0106:2,0:53:1:7.7:1:0.11208:0:60:0:0:0:3.7

[PolinaBevad]

Hi @asmlgkj , Wow, that's strange... Do you have a minimal example to reproduce this?

[asmlgkj]

Thanks a lot, it has been a long time to see you again, is everything ok with you, all vardict users miss you aso much. I think it was mainly because the sample had too many variants, it has nearly 500, 000 variants, and I ran a long time without ending, then I killed the process, and ran the same command again, this maybe the reason, in fact the file also has the repeated line apart from incomplete line. I have used --nosv, and set -th 16. the input file is not 300M(umi tumor) vs 600M (dedup normal)

is there any measures to speed up?

[PolinaBevad]

Thanks! Everything is fine, just there are many other projects, sorry for not coming back for so long! :)

Yes, that can be a reason, if you killed the process and it gave out only a piece of string from one thread and the full variants from another thread (if BED regions intersect). Input files are not very big, seems that the samples are very noisy? Do they have a lot of soft-clips? Is mapping quality ok - probably you can change a threshold? Maybe you can limit panel size (BED) only to the specific regions? Realignment step also take time, but I wouldn't recommend to disable it if you want to keep long indels.

[asmlgkj]

thanks a lot, I am very glad to hear everything is fine. here is my argument of vardict: -f 0.001 --nosv -r 2 -Q 30 -th 16. I used the default arguments of vardict of quality, do you have suggestions about this

and here is my bwa arguments : mem -Y -M -t 36