What is X-Score?


The X-Score program is developed by Dr. Renxiao Wang in Dr. Shaomeng Wang's group at the Department of Internal Medicine, University of Michigan Medical School. Here the prefix "X" indicates that it is integrated in the computer-aided drug design toolkit, X-Tool, which is being developed by Dr. Renxiao Wang.

X-Score is a "scoring function", which has its major applications to structure-based drug design studies. It computes a binding score for a given protein-ligand complex structure, and this binding score correlates to experimental binding constants well. Three individual empirical scoring functions have been implemented in this program, which are named as HPScore, HMScore, and HSScore, respectively. They can be conceptually written as:

HPScore = C0,1 + CVDW,1*(VDW) + CHB,1*(H-Bond) + CHP*(Hydrophobic Pair) + CRT,1*(Rotor)

HMScore = C0,2 + CVDW,2*(VDW) + CHB,2*(H-Bond) + CHM*(Hydrophobic Match) + CRT,2*(Rotor)

HSScore = C0,3 + CVDW,3*(VDW) + CHB,3*(H-Bond) + CHS*(Hydrophobic surface) + CRT,3*(Rotor)

By default, the final results: X-Score = (HPScore + HMScore + HSScore) / 3. However, the users are allowed to choose any combination of the above three scoring functions to meet their own interests.

All of these scoring functions were originally calibrated with a set of 200 diverse protein-ligand complexes, and they were able to reproduce the binding constants of the entire set with an average accuracy of 1.5 log units (corresponding to ~ 2 kcal/mol in terms of binding free energy). For more detailed description of X-Score, please check the reference given below.

Since X-Score does not perform molecular docking by itself, it is typically applied in combination with a molecular docking program, such as DOCK, AutoDock, FlexX or GOLD, in structure-base drug design studies. The molecular docking program is used to provide the binding models of the molecules of interests to their target protein. Then, X-Score can be applied to re-rank them and give more accurate estimation of the binding free energies of these molecules.

The correct way to refer this program is X-Score. Note that X-Score was formerly known as X-CScore for a short while. Please stop using the name of X-CScore to avoid further confusion. The following reference is supposed to be cited in any resulting publication by applying X-Score:

Wang, R.; Lai, L.; Wang, S. Further Development and Validation of Empirical Scoring Functions for Structure-Based Binding Affinity Prediction. J. Comput.-Aided Mol. Des. 2002, 16, 11-26.


What is new in X-Score v1.2?

Compared to the previous version, X-Score version 1.2 has substantial improvements in many aspects:

  1. All of the three scoring functions implemented in X-Score have been re-calibrated with a large training set of 800 protein-ligand complexes. X-Score v1.2 is thus applicable to a wider range of biomolecular systems and gives more robust results.
  2. Users have greater controls now on the scoring functions used in computation. They are allowed to adjust the coefficients before each term in the scoring functions to emphasize or underscore certain type of interactions between the protein and the ligand molecules if this is necessary for the success of their projects.
  3. More detailed outputs are available, which tabulate the role of each atom in binding score computation. This information helps users to analyze the structure-affinity relationship of the ligand molecules under study in a more transparent manner.
  4. New utilities have been added to aid users to prepare the input structures more conveniently and more accurately.
  5. Some bugs found in the previous version have been fixed. Many parts of the program have been optimized to give better performance.
  6. The user manual has been largely rewritten to reflect the many changes in version 1.2. New examples have been provided in the X-Score package for users to practice with this program.

[Content] [Introduction] [Download] [Installation] [Usage] [Utilities] [Trouble Shooting] [References]