What counts as validation,
and what does not
Part I chose the epitopes and Part II built the construct. This part takes apart, one at a time, the seven things called checks.
Docking, normal modes, population coverage, immune simulation, codon optimisation, RNA folding, RNA-TLR docking. Not one of the seven was measured at a bench.
That does not make them useless — in-silico screening narrows the question at a speed no laboratory can match. The question here is narrower: what do we call the result?
Python runs right inside this page. There is nothing to install.
How you'll learn
ClusPro scores, C-IMMSIM antibody counts and HDOCK confidence have no units or internal ones. First ask whether they can be compared at all.
Twenty-two hydrogen bonds against six — and the score points the other way. The scale you pick decides the conclusion.
97.49% and 23.88% sit in the same table. Which one reaches the abstract is the claim.
Predicted, suggested, supported, showed, proved. Lesson 20 draws the line on that scale.