Call SVs by genome assemblies, and genotyped in WGS population

[QianghuiZhu]

Hello!
Thank you for this nice tool. I want to know if I can call SVs by genome assemblies instead of Manta, and then genotype SVs in the WGS population data. My processes are as following. Hope you can give me some suggestions.

1. Using Mummer nucmer to align two genomes A and B.
2. Using Asseblytics, SVMU, SyRI to detect SVs based on mapping results.
3. Convert the result files from step2 to vcf4.3 files by home scripts. It should be clarified that I use allele sequence in VCF REF and ALT column instead of "N" or "<DEL>" symbolic alleles.
4. Merge Asseblytics, SVMU, SyRI results by SURVIVOR.
5. Genotyping SV datasets in WGS population by graphtyper.
6. Merge SVs with bcftools concat following your Wiki(https://github.com/DecodeGenetics/graphtyper/wiki/User-guide).

While Step5, I only got some warnings without any ERROR like this:

I have noticed that it is not suggested to use SURVIVOR to merge SVs(#42). But if I reserve the inserted sequence in VCF ALT, may I use graphtyper directly? Or does inserted sequence must have to be included in INFO SVINSSEQ?

If there are some error steps, could you point out and give me some suggestions to call SVs and genotype in WGS population to contain more SV sites (call SVs only by WGS data may have higher FDR, and lose many SV sites).

With great thanks to you.