Detecting gene insertions with DiscoSNP

Question

[laeb]

Hi, We have 2 bacterial strains. Parent and child. We have inserted a gene in the child strain. When we run discoSNP on reseq data of those 2 strains it only finds a single SNP difference. We have mapped raw reads to the plasmid we used to insert the gene and verified that the gene is indeed present in the child and not the parent.

Shouldnt DiscoSNP notice a large piece missing from the parent strain compared to the child strain and call it?

Best, Lasse

[Brian Bushnell]

SNP callers are not suited for detection and reporting of large-scale structural variations. If it is a simple insertion in the child, though, you can call it as a deletion in the parent by mapping the parent's data to the child with an aligner capable of handling long deletions (such as BBMap) to get the exact coordinates of the deletion (calling deletions is much easier than calling insertions). It can then be called normally or seen in IGV. If it is a more complex event, try a SV-caller.

It's not clear to me whether you inserted a gene from a plasmid into the child's main chromosome, or inserted a plasmid into the child. If you simply inserted a plasmid, there would be no variation (in the substitution/insertion/deletion model) since the original chromosome is unaffected.