Not able to merge insertions

[beantkapoor786]

Hello Fritz,

We have a tree genome pedigree in which we are comparing 5 different genomes to the reference genome. We aligned each genome using mummer4 and called structural variants. Now, we want to merge those structural variants using SURVIVOR's merge function. We were able to merge deletions and inversions but cannot merge insertions. Is it because the start and end positions in insertions are in the format END=START+1 and SVLEN=1? Please provide your thoughts. I am using the following command to do the merge but output vcf file is empty.

SURVIVOR merge insertions.txt 10000 1 1 1 0 1 mummer_insertions_merged.vcf

Output message -

merging entries: 51273
merging entries: 79656
merging entries: 75335
merging entries: 77095
merging entries: 79632

I have also attached the insertion vcf files here in case you want to look at them. Thanks!
insertions.zip

[sd880131]

I have the same problem

[fritzsedlazeck]

Hey there,
mmh the SVLen =1 is an issue. I wrote that wrapper script a while back for something... are you using this converter script from SURVIVOR for mummer ?

Btw. I would suggest this: (10k is a bit big)
SURVIVOR merge insertions.txt 10000 1 1 0 0 1 mummer_insertions_merged.vcf

The strand information might also be causing the issue...

Thanks
Fritz

[JiaB-He]

Hey there, mmh the SVLen =1 is an issue. I wrote that wrapper script a while back for something... are you using this converter script from SURVIVOR for mummer ?

Btw. I would suggest this: (10k is a bit big) SURVIVOR merge insertions.txt 10000 1 1 0 0 1 mummer_insertions_merged.vcf

The strand information might also be causing the issue...

Thanks Fritz

Hello, I did not find this conversion script

[zhiyongli1995]

I have the same problem. @beantkapoor786 Have you solved the problem? If so, please share the method. Thank you!

[beantkapoor786]

Hello, apologies for the late response. I couldn't do it using survivor but found another tool to merge the SVs - https://github.com/DecodeGenetics/svimmer

Hope this helps.