Jacobs University Bremen, School of Engineering and Science, Campus Ring 1, D-28759 Bremen, Germany.
Expert Opin Drug Discov. 2010 Apr;5(4):347-59. doi: 10.1517/17460441003713462.
Current computational docking methods are often effective in predicting accurate drug-binding geometries in cases of relatively rigid target structures. However, binding of drug-like ligands to protein receptor molecules frequently involves or even requires conformational adaptation. Realistic prediction of ligand-receptor binding geometries and complex stability needs in many cases an appropriate inclusion of conformational changes, not only for the ligand, but also for the receptor molecule.
Recent approaches to efficiently account for target receptor flexibility during docking simulations are reviewed.
The reader will gain insights into methods to efficiently treat protein side-chain flexibility and approaches for continuous adaptation of backbone conformations in pre-calculated essential or soft collective degrees of freedom. In addition, molecular dynamics or Monte Carlo based methods providing simultaneous inclusion of receptor and ligand flexibility are discussed as well as promising new developments to generate conformationally diverse ensembles of a protein structure. The large variety of possible conformational changes in proteins on ligand binding is illustrated for the enzyme reverse transcriptase of HIV-1, which is an important drug target.
If the backbone conformation of the binding site does not change, current docking programs can perform well by taking side-chain reorientations into account only. Future progress to account for full target flexibility in docking requires both accurate prediction of the essential modes of backbone motion and improvements in scoring to enhance selectivity. Thus, the scoring function should realistically cover energetic and particularly entropic contributions to binding, which would allow more realistic estimates of binding free energies.
目前的计算对接方法在目标结构相对刚性的情况下，通常能有效地预测准确的药物结合几何形状。然而，药物样配体与蛋白受体分子的结合经常涉及甚至需要构象适应。在许多情况下，需要适当考虑构象变化，不仅针对配体，还针对受体分子，以实现配体-受体结合几何形状和复杂稳定性的真实预测。
本文回顾了在对接模拟过程中有效考虑靶受体柔性的最新方法。
读者将深入了解有效处理蛋白侧链柔性的方法，以及在预先计算的基本或软整体自由度中连续适应骨架构象的方法。此外，还讨论了基于分子动力学或蒙特卡罗的同时包含受体和配体柔性的方法，以及生成蛋白结构构象多样性集合的有前途的新方法。在 HIV-1 逆转录酶等重要药物靶标上，说明了蛋白结合配体时的大量可能构象变化。
如果结合位点的骨架构象不变，当前的对接程序可以通过仅考虑侧链重排来很好地执行。在对接中考虑完整靶标柔性的未来进展需要准确预测骨架运动的基本模式，并改进评分以提高选择性。因此，评分函数应真实地涵盖结合的能量和特别是熵贡献，从而可以更真实地估计结合自由能。
