Table 1 The summary of molecular docking methods’ unique features and scoring functions, used in the CoBDock
From: Cobdock: an accurate and practical machine learning-based consensus blind docking method
Docking Methods | Pose research method | Advantages |
|---|---|---|
Autodock Vina | An empirical scoring function that was largely inspired by x-score [33] | \(\bullet\) High performance: 81% accuracy [43] |
\(\bullet\) Fast | ||
\(\bullet\) Ease of use | ||
\(\bullet\) Most common | ||
\(\bullet\) A high number of different pose locations | ||
ZDOCK | Energy-based scoring function (IFACE Statistical Potential, Shape Complementarity, and Electrostatics)) [40] | \(\bullet\) High performance: 85.71% [44] |
\(\bullet\) Blind (Global) docking | ||
\(\bullet\) A high number of poses | ||
PLANTS | PLANTS(CHEMPLP) or PLANTS(PLP) derivied from piecewise linear potential (PLP) scoring function [36] | \(\bullet\) High performance: 87% accuracy for the Astex Diverse Set (ADS) [24] |
\(\bullet\) Relatively fast | ||
\(\bullet\) A high number of variables related to ligand pose | ||
GalaxyDock3 | Global optimisation of a designed score function trained with an additional bonded energy term [22] | \(\bullet\) High performance [23] |
\(\bullet\) Ease of use | ||
\(\bullet\) Flexibility | ||
\(\bullet\) A high number of poses | ||
\(\bullet\) A high number of different pose locations |