いくつかの既存の教師データとなる既知PYMOL結果を用いて

ベースとなるリード分子に付加的な分子設計を行い

認識パターン用の特徴ベクトル

データをつくって探査解析する

①まず、低分子と蛋白のAUTODOCK結果を得る。アンてかエネルギー値とRSMDを得る。

②PYMOLに移して、観察データをとる。ここでの特徴データのとり方が勝負ポイントで、例えば結合部の近傍のαヘリックスの数と回線数の積が

アロステリック効果に関係するかもしれない、とにかく既存のうまく機能してる薬剤を教師データ化する必要がある

③①②のデータを使って、主成分解析やEMアルゴリズム解析を行う

Using PyMOL as a platform for
computational drug design
Shuguang Yuan ,
1* H.C. Stephen Chan2 and Zhenquan Hu3
PyMOL, a cross-platform molecular graphics tool, has been widely used for
three-dimensional (3D) visualization of proteins, nucleic acids, small molecules,
electron densities, surfaces, and trajectories. It is also capable of editing molecules, ray tracing, and making movies. This Python-based software, alongside
many Python plugin tools, has been developed to enhance its utilities and facilitate the drug design in PyMOL. To gain an insightful view of useful drug design
tools and their functions in PyMOL, we present an extensive discussion on various molecular modeling modules in PyMOL, covering those for visualization and
analysis enhancement, protein–ligand modeling, molecular simulations, and drug
screening. This review provides an excellent introduction to present 3D structures
visualization and computational drug design in PyMOL. © 2017 John Wiley & Sons, Ltd
How to cite this article:
WIREs Comput Mol Sci 2017, e1298. doi: 10.1002/wcms.1298
INTRODUCTION
PyMOL is an open-source molecular visualization
system created by Warren Lyford DeLano and
commercialized initially by DeLano Scientific LLC. In
2010, Schrödinger Inc. reached an agreement to
acquire PyMOL. From then on, Schrödinger has
taken over the development and maintenance, as well
as support and sale of PyMOL, including all current
subscription. PyMOL uses OpenGL Extension Wrangler Library (GLEW) and Free OpenGL Utility
Toolkit (Freeglut). PyMOL uses cross-platform
widget toolkit (Tk) for the GUI widgets. PyMOL can
produce high-quality movies and images of macromolecules in different representations including ribbons, cartoons, dots, surfaces, spheres, sticks, and
lines (Figure 1). At present, PyMOL is one of the
most widely used macromolecular visualization tools.
Since PyMOL is written in Python, one of the
most popular programming languages, it can be
extended to Python plugins easily. Apart from discussing the visualization and the enhanced analysis
functions in PyMOL, our topics also extend to the
protein–ligand modeling, molecular simulations
(MS), and virtual screening (VS) unities in PyMOL.
The computational drug discovery function of
PyMOL has been successfully applied to find new
drug candidates for various targets. These include the
discovery of a potent small molecule inhibitor for
gankyrin,1 lead optimization for Cytochrome P450
enzymes,2 VS of new drug candidates for the tumor
suppressor protein P53,3 and so on.
In this Software Focus article, we cover the
usage of PyMOL and its plugins as a platform for
computational drug design. We have summarized
PyMOL plugins and their functions in Table 1