I am doing protein-protein interface design using RosettaScripts, and in the .xml file under the demo at:
ROSETTAPATH/demos/public/rosetta_scripts_pp-interfacedesign/scripts
It has the following line:
<Ddg name="ddG" scorefxn="talaris2014" threshold="-15" repeats="2"/> binding energy calculation; an average of two repeats is computed for better numerical accuracy
Does anybody know what the rationale for that -15 threshold value is? Or for any other value that people use? I'm not sure if it's alright for me to use that value or somehow change it to fit my particular project.
Category:
Post Situation:
Typically these threshold values are determined by looking at your positive controls. You look for structures/systems where you have a known good answer (e.g. decent binding) and look at what sort of energy values you get for "decent" (though not necessarily perfect) binding. You can then set the threshold here such that your "positives" would pass, but the obvious negatives wouldn't.
The other approach is to forgo the actual filtering (e.g. set it really high, or switch to a non-filtering system like SimpleMetrics) and keep everything. If the ddG value gets reported in the scorefile, you can do a post hoc filtering based on the distribution of ddG values you see in practice.