Mykrobe overcounts homopol deletions when making probes #148
Description
The probe generator is making one probe for each 1bp and 2bp deletions in katG and pncA. But if you make a all 1bp deletions in a homopolymer, they are essentially the same.
eg deleting one A from AAA could happen in 3 places and give the same sequence.
This leads to annoying things where we report >1 mutation being detected, when it is basicvally the same thing detected twice/whatever.
Workaround suggestions
- compare all the indel/frame-shift probes against each other and dedup
- pay attention to whether you are in a homopolymer when creating the 1bp and 2bp deletions, and don't overcreate probes