Datar Prasanna A, Khedkar Santosh A, Malde Alpeshkumar K, Coutinho Evans C
Department of Pharmaceutical Chemistry, Bombay College of Pharmacy, Kalina, Santacruz (E), Mumbai 400098, India.
J Comput Aided Mol Des. 2006 Jun;20(6):343-60. doi: 10.1007/s10822-006-9051-5. Epub 2006 Sep 29.
A novel approach termed comparative residue-interaction analysis (CoRIA), emphasizing the trends and principles of QSAR in a ligand-receptor environment has been developed to analyze and predict the binding affinity of enzyme inhibitors. To test this new approach, a training set of 36 COX-2 inhibitors belonging to nine families was selected. The putative binding (bioactive) conformations of inhibitors in the COX-2 active site were searched using the program DOCK. The docked configurations were further refined by a combination of Monte Carlo and simulated annealing methods with the Affinity program. The non-bonded interaction energies of the inhibitors with the individual amino acid residues in the active site were then computed. These interaction energies, plus specific terms describing the thermodynamics of ligand-enzyme binding, were correlated to the biological activity with G/PLS. The various QSAR models obtained were validated internally by cross validation and boot strapping, and externally using a test set of 13 molecules. The QSAR models developed on the CoRIA formalism were robust with good r (2), q (2) and r (pred) (2) values. The major highlights of the method are: adaptation of the QSAR formalism in a receptor setting to answer both the type (qualitative) and the extent (quantitative) of ligand-receptor binding, and use of descriptors that account for the complete thermodynamics of the ligand-receptor binding. The CoRIA approach can be used to identify crucial interactions of inhibitors with the enzyme at the residue level, which can be gainfully exploited in optimizing the inhibitory activity of ligands. Furthermore, it can be used with advantage to guide point mutation studies. As regards the COX-2 dataset, the CoRIA approach shows that improving Coulombic interaction with Pro528 and reducing van der Waals interaction with Tyr385 will improve the binding affinity of inhibitors.
一种名为比较残基相互作用分析（CoRIA）的新方法已被开发出来，该方法强调在配体 - 受体环境中定量构效关系（QSAR）的趋势和原理，用于分析和预测酶抑制剂的结合亲和力。为了测试这种新方法，选择了一个由九个家族的36种COX - 2抑制剂组成的训练集。使用DOCK程序搜索抑制剂在COX - 2活性位点的假定结合（生物活性）构象。对接构型通过蒙特卡罗方法和模拟退火方法与Affinity程序相结合进一步优化。然后计算抑制剂与活性位点中各个氨基酸残基的非键相互作用能。这些相互作用能，加上描述配体 - 酶结合热力学的特定项，通过广义偏最小二乘法（G/PLS）与生物活性相关联。获得的各种QSAR模型通过内部交叉验证和自助法进行验证，并使用一组13个分子的测试集进行外部验证。基于CoRIA形式主义开发的QSAR模型具有稳健性，r (2)、q (2) 和r (pred) (2) 值良好。该方法的主要亮点包括：在受体环境中采用QSAR形式主义来回答配体 - 受体结合的类型（定性）和程度（定量），以及使用能够解释配体 - 受体结合完整热力学的描述符。CoRIA方法可用于在残基水平上识别抑制剂与酶的关键相互作用，这在优化配体的抑制活性方面可得到有效利用。此外，它还可有利地用于指导点突变研究。对于COX - 2数据集，CoRIA方法表明，提高与Pro528的库仑相互作用并减少与Tyr385的范德华相互作用将提高抑制剂的结合亲和力。
