| Attributs | Valeurs |
|---|
| type
| |
| Is Part Of
| |
| Subject
| |
| Title
| - AIScore Chemically Diverse Empirical ScoringFunction Employing Quantum Chemical Binding Energies of Hydrogen-BondedComplexes
|
| has manifestation of work
| |
| related by
| |
| Author
| |
| Abstract
| - In this work we report on a novel scoring function that is basedon the LUDI model and focuses on the prediction of binding affinities.AIScore extends the original FlexX scoring function using a chemicallydiverse set of hydrogen-bonded interactions derived from extensivequantum chemical ab initio calculations. Furthermore,we introduce an algorithmic extension for the treatment of multifurcatedhydrogen bonds (XFurcate). Charged and resonance-assisted hydrogenbond energies and hydrophobic interactions as well as a scaling factorfor implicit solvation were fitted to experimental data. To this end,we assembled a set of 101 protein−ligand complexes with knownexperimental binding affinities. Tightly bound water molecules inthe active site were considered to be an integral part of the bindingpocket. Compared to the original FlexX scoring function, AIScore significantlyimproves the prediction of the binding free energies of the complexesin their native crystal structures. In combination with XFurcate,AIScore yields a Pearson correlation coefficient of RP = 0.87 on the training set. In a validationrun on the PDBbind test set we achieved an RP value of 0.46 for 799 attractively scoredcomplexes, compared to a value of RP = 0.17 and 739 bound complexes obtained with the FlexX originalscoring function. The redocking capability of AIScore, on the otherhand, does not fully reach the good performance of the original FlexXscoring function. This finding suggests that AIScore should ratherbe used for postscoring in combination with the standard FlexX incrementalligand construction scheme.
|
| Alternative Title
| - AIScore Chemically Diverse Empirical ScoringFunction
|
| is part of this journal
| |
