National Supercomputer Center in Guangzhou, School of Data and Computer Science, Sun Yat-sen University, 132 East Circle at University City, Guangzhou, 510006, China
Auto-generated profile, based on publicly available information
Jiahui Li
National Supercomputer Center in Guangzhou, School of Data and Computer Science, Sun Yat-sen University, 132 East Circle at University City, Guangzhou, 510006, China
Performance of structure-based molecular docking largely depends on the accuracy of scoring functions. One important type of scoring functions are knowledge-based potentials derived from known three-dimensional structures of proteins and/or protein–ligand complex structures. This study seeks to improve a knowledge-based protein–ligand potential based on a distance-scale finite ideal-gas reference (DFIRE) state (DLIGAND) by expanding the representation of protein atoms from 13 mol2 atom types to 167 residue-specific atom types, and employing a recently updated dataset containing 12,450 monomer protein chains for training. We found that the updated version DLIGAND2 has a consistent improvement over DLIGAND in predicting binding affinities for either native complex structures or docking-generated poses. More importantly, DLIGAND2 has a 52% increase over DLIGAND in enrichment factors in top 1% predictions based on the DUD-E decoy set, and consistently improves over Autodock Vina and other statistical energy functions in all three benchmark tests. We further found that DLIGAND2 outperforms empirical and machine-learning methods compared for virtual screening on new targets that are not homologous to the DUD-E training set. Given the best performance as a parameter-free statistical potential and among the best in all performance measures, DLIGAND2 should be useful for re-assessing the poses generated by docking software, or acting as one term in other scoring functions. The program is available at https://github.com/sysu-yanglab/DLIGAND2 .
Background: The increasing number of organic and inorganic structures promotes the development of the “Big Data” in chemistry and material science, and raises the need for cross-platform and web-based methods to search, view and edit structures. Many web-based three-dimensional (3D) structure tools have been developed for displaying existing models, building new models, and preparing initial input files for external calculations. But few of these tools can deal with crystal structures. Results: We developed a user-friendly and versatile program based on standard web techniques, such as Hyper Text Markup Language 5 (HTML5), Cascade Style Sheet (CSS) and JavaScript. Both non-periodic organic molecule and crystal structure can be visualized, built and edited interactively. The atom, bond, angle and dihedral in a molecule can be viewed and modified using sample mouse operations. A wide range of cheminformatics algorithms for crystal structure are provided, including cleaving surfaces, establishing vacuum layers, and building supercells. Four displayed styles, namely “Primitive cell”, “Original”, “In-cell” and “Packing” can be used to visualize a unit cell. Additionally, the initial input files for Vienna Ab-initio Simulation Package (VASP) and Gaussian can be obtained by interacting with dialog boxes in 3DStructGen. Conclusions: 3DStructGen is a highly platform-independent program. It can provide web service independently or can be integrated into other web platforms. Other than local desktop software, it does not require any additional effort to install the system but a web browser supporting HTML5. 3DStructGen may play a valuable role in online chemistry education and pre-processing of quantum calculations. The program has been released under MIT open-source license and is available on: https://matgen.nscc-gz.cn/Tools.html.