Why some insertions are symbolic in ALT column?

Question

[Maxine]

I have a VCF file generated using the SV caller Sniffles2, and I have noticed that some ALTs for insertions are represented symbolically as "<INS>" instead of providing the actual variant sequence in the ALT column.

Initially, I believed that these insertions were represented symbolically due to their length exceeding a certain threshold. To investigate this, I conducted some statistics. However, the results did not support my initial assumption.

The length of symbolic insertions ranges from 2341 to 78565, while the length of non-symbolic insertions ranges from 31 to 49858.

I am curious to know what factors determine whether an insertion is represented symbolically in a VCF file. And, is it possible to obtain the actual variant sequence of symbolic insertions?

Answer 1

[LChart]

I don't know specifically about Sniffles2; but for other callers this typically means that the insert sequence itself couldn't be assembled or is ambiguous.

[Medhat]

I would like to confirm what was just mentioned by LChart (I am one of Sniffles2 developers)

[Maxine]

sounds great! waiting for your findings~

[Maxine]

thank you for sharing your experience!

Answer 2

[Qi]

i also meet this problem now, after reading your post, I decided to remove these variations

[Maxine]

yes, I also removed them from the vcf. Just for referrence, Fritz, the designer of sniffles2, said it is because the SV sequences are different among samples. See it at this link.

[Qi]

thanks! it is a excellent tools to identify sv.