Similarity thresholds
interaction: 0.900,
known: 0.950
Rule bonus
all-models-pass: 15,
interaction: 10,
known: 5
Rule Bonus Raw =
(Bonus all-models-pass × AllModelsPassDocking)
+ (Bonus interaction × PassSimilarityInteraction)
+ (Bonus known × PassSimilarityKnown)
Rule Bonus Norm = Rule Bonus Raw ÷
(Bonus all-models-pass + Bonus interaction + Bonus known)
Priority weights
docking: 40,
similarity: 30,
consistency: 20,
ruleBonus: 10
Priority Score =
(Docking Weight × Docking Mean Score)
+ (Similarity Weight × Similarity Score Norm)
+ (Consistency Weight × Consistency Score)
+ (Rule Bonus Weight × Rule Bonus Norm)
Docking Mean Score usesreverse min-max normalization within the current scorable batch;
Similarity Score Normuses forward min-max normalization;
Consistency Score uses reverse min-max normalization on Docking Std.
Higher Priority Score means higher relative priority within the current batch; it does not by itself mean that the candidate has passed the platform threshold criteria.