Improving ligand‐ranking of AutoDock Vina by changing the empirical parameters
T Ngoc Han Pham,Trung Hai Nguyen,Nguyen Minh Tam,Thien Y Vu,Nhat Truong Pham,Nguyen Truong Huy,Binh Khanh Mai,Nguyen Thanh Tung,Minh Quan Pham,Van V Vu,Son Tung Ngo,T. Ngoc Han Pham,Thien Y. Vu,Van V. Vu
DOI: https://doi.org/10.1002/jcc.26779
2021-10-30
Journal of Computational Chemistry
Abstract:AutoDock Vina (Vina) achieved a very high docking-success rate, <math><mover><mi>p</mi><mo>^</mo></mover></math> , but give a rather low correlation coefficient, <math><mi>R</mi></math> , for binding affinity with respect to experiments. This low correlation can be an obstacle for ranking of ligand-binding affinity, which is the main objective of docking simulations. In this context, we evaluated the dependence of Vina R coefficient upon its empirical parameters. <math><mi>R</mi></math> is affected more by changing the gauss2 and rotation than other terms. The docking-success rate <math><mover><mi>p</mi><mo>^</mo></mover></math> is sensitive to the alterations of the gauss1, gauss2, repulsion, and hydrogen bond parameters. Based on our benchmarks, the parameter set1 has been suggested to be the most optimal. The testing study over 800 complexes indicated that the modified Vina provided higher correlation with experiment <math><msub><mi>R</mi><mrow><mi>set</mi><mn>1</mn></mrow></msub><mo>=</mo><mn>0.556</mn><mo>±</mo><mn>0.025</mn></math> compared with <math><msub><mi>R</mi><mtext>Default</mtext></msub><mo>=</mo><mn>0.493</mn><mo>±</mo><mn>0.028</mn></math> obtained by the original Vina and <math><msub><mi>R</mi><mrow><mtext>Vina</mtext><mspace></mspace><mn>1.2</mn></mrow></msub><mo>=</mo><mn>0.503</mn><mo>±</mo><mn>0.029</mn></math> by Vina version 1.2. Besides, the modified Vina can be also applied more widely, giving <math><mi>R</mi><mo>≥</mo><mn>0.500</mn></math> for 32/48 targets, compared with the default package, giving <math><mi>R</mi><mo>≥</mo><mn>0.500</mn></math> for 31/48 targets. In addition, validation calculations for 1036 complexes obtained from version 2019 of PDBbind refined structures showed that the set1 of parameters gave higher correlation coefficient ( <math><msub><mi>R</mi><mrow><mi>set</mi><mn>1</mn></mrow></msub><mo>=</mo><mn>0.617</mn><mo>±</mo><mn>0.017</mn></math> ) than the default package ( <math><msub><mi>R</mi><mi>Default</mi></msub><mo>=</mo><mn>0.543</mn><mo>±</mo><mn>0.020</mn></math> ) and Vina version 1.2 ( <math><msub><mi>R</mi><mrow><mtext>Vina</mtext><mspace></mspace><mn>1.2</mn></mrow></msub><mo>=</mo><mn>0.540</mn><mo>±</mo><mn>0.020</mn></math> ). The version of Vina with set1 of parameters can be downloaded at https://github.com/sontungngo/mvina. The outcomes would enhance the ranking of ligand-binding affinity using Autodock Vina.
chemistry, multidisciplinary