Sniffles2 output

[annecmg]

Hello, I am new to structural variant calling and I am looking for a bit more information on how to interpret the Sniffles2 output. I have used multi-sample calling on a haploid genomes and now I am unsure how to interpret in which samples an SV was found, e.g.,

Fol007_1 7177 Sniffles2.INS.0M0 N TGGTTCATAGACTTTTTGTTTGTTTGTTTGTTTGTTTGTTTGTATATTTTTCGTCTGGTGCTGTAACGAAGCCAGATCTCTACTGGAGCAAAAAAAGGACGTGCTGAGCTAATAGGTACAGGGAAACCCCCGCTTCCTTCACCATCCAGCATACCAATGGCACAAAGTGCTGTATTTCTCAAGCCCGTCACCCGTCAGCGGGTTCAGGGGCAATCTATTGTCGAACTTTTTCCACCGACTGCAGAATTCTCTCGCGCATCACTACATCTGCCGACGCAGTCCACAATCTCCAATGGCAAGTTGTAGCTCGTAGTCGACAATGTCGCGGCGGCAGGGAAAAACCTCCGGGAAGAATCAGAGGCCGACGGTGCGGCTTGCCTTCAAGGAGAAAAAGCCCATCGCAATCACGGCGTCGTTGATCAGGGGATGTCGCCACCGATGCCATGGGCATCGTCCTCCGATATCTAGGCGTTCTGGTTCTGAGGTGCCGCCATTAGATCCGACAAGGTGTTGTCGAAGGCAGACTCTCCCAGAATTCCTTCTCTGAAAACTGCTGTTGGGGAACCGTGTATCCGCAGTTTAATGTGTCGCGTGACTGAAAAATCCACTCCTAACACGGCACTAACAGTGCGTGCCAAGAAAAGAATTTTTTACGGGGAGGAGTTG 53 PASS PRECISE;SVTYPE=INS;SVLEN=670;END=7177;SUPPORT=3;COVERAGE=14,14,14,15,15;STRAND=+-;AC=2;STDEV_LEN=11.719;STDEV_POS=8.083;SUPP_VEC=001000000001100000000000000000000000000 GT:GQ:DR:DV:ID ./.:0:3:0:NULL ./.:0:1:0:NULL 0/0:45:20:1:Sniffles2.INS.0S0 0/0:0:18:0:NULL 0/0:0:10:0:NULL 0/0:0:10:0:NULL 0/0:0:8:0:NULL 0/0:0:5:0:NULL 0/0:0:11:0:NULL 0/0:0:10:0:NULL ./.:0:2:0:NULL 1/1:6:1:6:Sniffles2.INS.0S0 0/0:29:14:1:Sniffles2.INS.0S0 ./.:0:2:0:NULL ./.:0:4:0:NULL ./.:0:3:0:NULL 0/0:0:11:0:NULL ./.:0:0:0:NULL ./.:0:0:0:NULL 0/0:0:24:0:NULL 0/0:0:19:0:NULL 0/0:0:22:0:NULL 0/0:0:18:0:NULL 0/0:0:14:0:NULL ./.:0:2:0:NULL ./.:0:4:0:NULL 0/0:0:5:0:NULL 0/0:0:10:0:NULL ./.:0:0:0:NULL ./.:0:1:0:NULL 0/0:0:22:0:NULL 0/0:0:22:0:NULL ./.:0:4:0:NULL 0/0:0:23:0:NULL 0/0:0:5:0:NULL 0/0:0:6:0:NULL ./.:0:0:0:NULL ./.:0:0:0:NULL ./.:0:0:0:NULL

According to SUPP_VEC this is found in the 3rd sample, but that is genotyped as 0/0. Although there is only 1 reference read, it corresponds to an insertion in the original call (0/0:45:20:1:Sniffles2.INS.0S0). Should this be interpreted as having the SV present or shall only the ones genotyped as 0/1 or 1/1 be included?
Thanks a lot, Anne

[fritzsedlazeck]

Dear Anne,
so I know this is confusing.. we are using 0/0 but there might be evidence on the read level. In your case, I am assuming you sequenced a population of individuals? So the read ration should be correlated with heterogenity in the population.... or of course could be also artifacts. So 0/0 doesnt always mean totally absent. The SUP_VEC indicates a 1 if a single read can be found and assigned to the SV.

hope that helps a bit ??
Bis bald :)
Fritz

[annecmg]

Thanks for the quick reply! As I understand one read would already be evidence, especially when the SV was also found in other samples? The reads come indeed from a population!

[fritzsedlazeck]

Correct. The merge step rescues other samples that might have too little evidence.. so if there is one passing call in one sample.. the others might get rescued.
Cheers
Fritz