search

ktmeaton / ncov-recombinant

Reproducible workflow for SARS-CoV-2 recombinant sequence detection.
MIT License
18 stars 2 forks source link

False positives caused by sc2rf XAJ mode #191

Closed ktmeaton closed 2 years ago

ktmeaton commented 2 years ago

There is an interesting phenomenon, where if you only search for recombination involvingBA.2/BA.5 you will get BA.4 lineages. And if you only search for recombination involving BA.2/BA.4, you will get BA.5 lineages.

The sc2rf mode to detect XAJ, searches for recombination between BA.4 and BA.2.12. Which picks up some BA.5* lineages as false positives.

Some options for making this more restrictive would be test to using additional clades, restricting breakpoints, increase unique subs, etc.

ktmeaton commented 2 years ago

I've chosen to increase unique subs --unique 2 and be really restrictive on breakpoints --breakpoints 2.

ktmeaton commented 2 years ago

This was implemented in #192 .