Menu
Home
Log in / Register
 
Home arrow Computer Science arrow Computational systems pharmacology and toxicology
Source

Scoring Ligand Poses

All docking algorithms employ scoring functions to determine the energetic favorability of any given ligand pose. In general, the equations used to calculate scoring functions encapsulate key factors that mediate ligand-receptor interactions (such as hydrogen bonds, solvation effects, polar interactions, and non-polar interactions). Overall, there are three major classes of scoring functions:13 (1) force-field based functions; (2) empirical functions; and (3) knowledge-based functions. In this section, we briefly describe the force- field and empirical scoring functions that are used by popular docking algorithms such as AutoDock, CHARMM and GOLD.

The docking algorithm in CHARMM14 employs a force-field based scoring function that uses Lennard-Jones potentials to assess van der Waals interactions between the ligand and its receptor while also summing the effects of electrostatic interactions between the ligand and receptor in the ligand-receptor complex. Overall, the major drawback to the use of a force- field based scoring function is the presence of local minima that may skew the resulting energy calculations and which, in turn, necessitates the use of energy minimization steps prior to scoring a ligand pose.

In contrast, AutoDock,2,1S AutoDock Vina6 and GOLD16 use semi-empirical or empirical scoring functions. Broadly speaking, these algorithms apply a weighting scheme to each interaction factor involved in describing the binding of a ligand to a receptor. These weights are calculated via analysis of a training data set and involve fitting the components of the scoring function to the experimentally derived binding constants. When using empirical or semi-empirical scoring functions, care must be taken to ensure that the ligand of interest is not overly dissimilar to the ligands used in the training set used to weight the various interaction terms.

Due to the inherent flaws in any single scoring function, alternate techniques can be used to reduce the likelihood of false-positives. These involve the use of consensus approaches where two or more scoring functions are used to independently rate any given pose.17 In these analyses, only poses that are deemed favorable by two or more scoring functions are taken to signify favorable energetics of interaction.

 
Source
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Mathematics
Political science
Philosophy
Psychology
Religion
Sociology
Travel