[chr6:46679539-46682097][DEL][SVLEN=2558bp]

[jzook]

chr6 46679539 Minda_117 N <DEL> . PASS SVLEN=2558;SVTYPE=DEL;SUPP_VEC=PB_d_381,ONT_ID_66217_2,ONT_d_166,PB_ID_42022_2,ONT_severus_DEL7093,ONT_Sniffles2.DEL.1CAM5,PB_severus_DEL6141,PB_Sniffles2.DEL.1A4M5,ILL_MantaDEL:114155:0:1:0:0:0,ILL_3138274171:2,ILL_gridss110fb_16480h

https://v2.genomeribbon.com/?session=https://42basepairs.com/download/s3/giab-data/ribbon-json/ribbon-hg008-cov-bedpe.json&locus=chr6:46679539#ribbon

[cbrueffer]

True homozygous deletion:

• Loss of coverage supported in Ribbon, but loss region is too small to be really visible in SplitThreader
• Breakpoints generally supported in Ribbon
• Left exact breakpoint supported in IGV
• Right breakpoints slight disagreement between ONT (exact) and PacBio (coordinate + 1) in IGV and thus disagreement about exact SVLEN.
• Left breakpoints within a SINE, right breakpoints right between two LINEs
• Deletion fully covers 3 SINEs and 2 LINEs

Looking at the SNP PacBio calls after the deletion (see last IGV screenshot) I'm more inclined to believe ONT (although that shows no coverage at the first "SNP" position?), but maybe someone more experienced with both technologies can weigh in.

[jzook]

Looks like different alignments of the same complex variant in HiFi and ONT. Seems like your IGV is set to ignore small indels in ONT, since there is 1bp with no coverage, so it is a net 2559bp del in both with nearby SNVs that make it complex

[cbrueffer]

Thanks for chiming in; very small indels were indeed set to be hidden. The IGV option is View -> Preferences -> Third Gen -> Hide indels < indel threshold for anyone watching.