How to interpret deletion the length of chromosome?

Question

[optimistsso4co3]

I called structural variants with MANTA, merged with SURVIVOR for 600 samples. And get thousands of Deletions that basically span the entire chromosome and even have high frequency. There are obviously other variants for these specific individuals for specific chromosome, so i wonder how is that possible?

Here is an example of VCF line for such variant: After manta:

chr1    789563  MantaDEL:22:0:2:1:0:0   **A       DEL**   16      MinQUAL END=224012356;SVTYPE=DEL;**SVLEN=-223222793**;IMPRECISE;CIPOS=-74,74;CIEND=-79,79   GT:FT:GQ:PL:PR  0/1:PASS:16:66,0,83:6,4

And after merging with SURVIVOR:

chr1    789563  MantaDEL:24:0:1:0:0:0   **C       DEL**   117     PASS    SUPP=48;SUPP_VEC=00[....]000;**SVLEN=-223222860**;SVTYPE=DEL;SVMETHOD=SURVIVOR1.0.7;CHR2=chr1;END=224012378;CIPOS=-77,74;CIEND=-55,1;STRANDS=+-        GT:PSV:LN:DR:ST:QV:TY:ID:RAL:AAL:CO    0/1:NA:223222874:0,0:+-:20:DEL:MantaDEL_33_0_1_0_0_0:NA:NA:chr1_789495-chr1_224012369 [........]

[Pierre Lindenbaum]

this is quite common , and this is a false positive

[optimistsso4co3]

Is there a way to remove false positives?

[Pierre Lindenbaum]

quick:

bcftools view -e 'SVLEN > 10000000 || SVLEN < -10000000'