could I know why set "--srv No" for busted model, thanks

veg / selection-tutorial

Creative Commons Zero v1.0 Universal

3 stars 1 forks source link

could I know why set "--srv No" for busted model, thanks #4

Closed liamxg closed 11 months ago

liamxg commented 11 months ago

hyphy busted --srv No --alignment data/HIV-sets.nex --starting-points 5 --output results/HIV-sets.nex.BUSTED.json

stevenweaver commented 11 months ago

No is the default BUSTED as specified in the original paper. SRV is an extension that accounts for synonymous rate variation specified in a later paper.

liamxg commented 11 months ago

@stevenweaver If No is the default, why you emphasize it in the command line?

stevenweaver commented 11 months ago

Not sure. I didn't write it.

liamxg commented 11 months ago

@stevenweaver the default is Yes, not No:

liamxg commented 11 months ago

Any which situation can use Yes?

stevenweaver commented 11 months ago

Yes is a more conservative test that accounts for synonymous rate variation. It would be difficult to expand much more beyond the Discussion in the BUSTED SRV paper without a more specific question regarding it.

stevenweaver commented 11 months ago

The comment powerfully recommends including SRV in your analyses.

The empirical and simulation work in this paper adds to a growing body of literature strongly suggesting that the models underlying these methods—not only BUSTED, but almost certainly any positive selection method that assumes the absence of SRV—are importantly wrong and need to be revised.

liamxg commented 11 months ago

@stevenweaver thanks again. I am much clear now. :)