Molecular Docking Simulation

1. Molecular structure files: Protein-ligand complex files for re-docking experiments were obtained from the PDBbind database. To validate predictive models with less bias, native ligands of the co-crystallized complexes were first extracted and converted into 2D using Open Babel [43] (link). For the following docking simulation, 2D structures were then re-converted to 3D using a 3D structure generator called CORINA version 3.4 [44] .
2. Molecular docking simulation packages: Native ligands were docked to their corresponding target proteins using eHiTS, GOLD, and AutoDock VINA (Table S7). These docking tools are used to generate numerous binding modes of the test compound in a defined binding site, and the number of binding modes generated varies with the docking tools. For a docking simulation, eHiTS was set to output 1000 conformations for each docking study. Considering the computing speed of GOLD, we set the maximum as 300. The maximum binding mode of AutoDock VINA varies with an energy range of 10 (kcal/mol).
3. Application of machine learning systems: Binding modes generated by the three docking tools were re-scored by machine learning system A, and only the three top-score candidates in each set were retained. Subsequently, machine learning system B assessed the three top-score candidates and identified the most predictive one. Modeling exercises of the machine learning systems A and B were conducted using the R statistical package. The Random Forest algorithm was applied to build machine learning system A, which was implemented in “randomForest” (Breiman and Cutler's random forests for classification and regression) module. For machine learning system B, the multinomial logistic regression of “nnet” (Feed-forward Neural Networks and Multinomial Log-Linear Models) and “MASS” (Modern Applied Statistics with S. Fourth Edition) modules was utilized.
4. Re-docking result: The Pearson correlation coefficient between the predicted docking scores and the experimental binding affinities was calculated using R to determine the predictiveness of the screening approach.

1. Protein structure files: protein structures collected from the PDB database complied with the following criteria: 1) X-ray structures with resolution of 2.5Å or better, if available 2) if two or more structures were available, that with the best solution was selected 3) a structure with a ligand bound to its nucleotide binding site was selected 4) non-modified and non-phosphorylated residues found in the binding site were selected with priority 5) the organism was human.
2. Test compound files: test compound structure files in 2D format were downloaded from PubChem, and converted into 3D using CORINA version 3.4 for the docking simulation.
3. Molecular docking simulation: the use of the docking tools was the same as mentioned in the re-docking experiment.

To explore the relationship and action mechanisms between candidate proteins and active ingredients, molecular docking simulations were conducted to evaluate the strength and mode of interactions between components and hub targets. Crystal structures of critical targets protein receptors were acquired from the Protein Data Bank database (http://www.rcsb.org/) in Protein Data Bank format. The active component structure as ligands was downloaded from the PubChem compound database (http://pubchem.ncbi.nlm.nih.gov/). After removing the water molecules and organic compounds from ligands and proteins and adding non-polar hydrogen bridge to them by PyMol 2.6.0 software, the format of the molecular ligands and proteins was transformed into pdbqt format. Subsequently, the docking of ligands and proteins was performed by AutoDockTools 1.5.7 software. Each group of molecular docking was run 50 times, and the ionization energy was calculated. The minimum energy value was selected as the docking affinity. Finally, the docking results were visualized using PyMol software.

We used MOE software to screen FDA-approved drugs that bind to target proteins and perform molecular docking simulations. Protein structures of core targets were collected from the PDB database and FDA-approved drugs were collected from the zinc15 database and converted to 3D structures in MOE by energy minimization. We optimized the protonation state of the protein and the direction of hydrogen at the PH of LigX 7 and the temperature of 300K. Finally, we studied the binding mode of PPARG, SLC7A9 and GALK1 with small molecule drugs by rigid docking simulation.