Sep 13 2024
Feedback.
|
The results table will contain the columns from the chemical table that you screened plus the following results columns:
Depending on your selection you might have two tables Model Pairwise and Target Pairwise
- mol - your query chemical
- target - the target model you screened against
- pPValue - kcc/hybrid4D+2D models only: -Log(probability) relative to approved drug decoys capped at 5: 1: Top 10 percent 2: Top 1 percent 3: Top 0.1 percent
- MolpKd - kcc/hybrid4D+2D models only: Prediction of pKd, pKi, pIC50, pEC50, etc >6. indicates submicromolar value, >9. indicates nanomolar, ND indicates disimilarity to known binders
- MolpKd Error - kcc/hybrid4D+2D models only: Estimated total uncertainty/error for pKd prediction, including experimental error of ~1. pKd unit
- MolClass - kcc/hybrid4D+2D models only: Classification Score: for kcc models: nProbability of being a positive, assume 0.5% hitrate for others: 1. Median value of actives 2.Median value of inactives
- TANIsim - kcc models only: Tanimoto distance from model nearest compound
- modelnearestCpd - kcc models only: nearest compound to the query in the model
- MolZScore - apf models only: Z-Score 3. indicates 3 standard deviation above approved drugs decoy
- predictType - model type
- modelAUC - Area Under Curve (AUC) for model
- modelpKdAUC - kcc models only: AUC for pKd model
- modelRMSE - kcc models only: RMSE for model
- modelQ2 - kcc models only: Q2 for model
- MolDockScore hybrid docking only - ICM docking score, Score < -32. are in general significant, but may vary from structure to structure.
Fold Columns to Original Table
In the Tools panel of the table there is a convenient way of packaging your results into your original query table.
|