Oxford Nanopore insertion vs deletion rates

Question

[BioinformaticsLad]

Indels are a big problem in Oxford Nanopore reads due to difficulty in basecalling homopolymers. Are insertions and deletions equally likely? If the one type is more likely than the other, why don't they compensate for it in the basecaller by systematically under- or over-calling homopolymer length?

[Vitis]

Based on my (limited) experience with Nanopore data, looks like there was no bias towards either insertions or deletions. I don't have hard data to support this, though.

Answer 1

[WouterDeCoster]

Short insertions and short deletions are about equally likely.