Can I turn off the "pathogenic/benign SV annotation" mode?

[lgmgeo]

Can I turn off the "pathogenic/benign SV annotation" mode?

[lgmgeo]

Hi fjmuzengyiheng,

In versions lower than v3.0, AnnotSV was reporting benign genomic regions partially overlapping the SV to annotate (>70%).
Now that I have a better understanding of the ACMG/ClinGen guidelines, I really believe that this criterion was not good enough to be used to filter out SV.
The idea of AnnotSV 3.0 was to stick to the new ACMG/ClinGen criteria (based in part on benign/pathogenic SV). AnnotSV searches for benign genomic regions completely overlapping the SV to annotate.

The "frequencies" from DDD, DGV, gnomAD, 1000g... are no longer reported with AnnotSV 3.0.
It's basically because the output was getting bigger and bigger (with the SV sources increase).

Actually, in v3.0, the frequencies has been used to create the benign/pathogenic SV data sources in AnnotSV.
For example, benign SV have been merged in AnnotSV from gnomAD, ClinVar, ClinGen, DGV, DDD, 1000g and IMH. The benign SV extraction method from each source is described in the README. As an example, putatively benign variants were selected from gnomAD i) if at least one population allele frequency > 1% OR ii) if the homozygous individuals count > 5.

So if you are interested in filtering SV from the frequencies, this is no longer possible...
The idea is to directly filter your SV with the “AnnotSV_ranking_score" and "ACMG_class" values (e.g. you can filter out the SV when ACMG_class = "1").

Best,
Véronique

[fjmuzengyiheng]

Thank you so much for explaining the detail to me.
Now I understand the efforts you update AnnotSV to alive version 3.0.
I decide to use 3.0 for my pipeline.
Thank you again for this outstanding tool.

[lgmgeo]

Happy users make me happy ;o)