Quantitative Structure-Activity Relationship, Activity Prediction, and Molecular Dynamics of Non-nucleotide Reverse Transcriptase Inhibitors

The research focuses on computer-aided drug design to develop new innovative drugs or repurpose currently existing drugs for African-related diseases. Okay, so in our field currently, we have a lot of machine learning applications that people are developing, and this is something that we also wish to incorporate within our group in order to design the beta drugs. Currently we make use of what's known as high-performance computing, and this is basically what speeds up our research by making use of both CPU and GPU technology in order to get results faster when we experiment uses.

At the moment, computational is still a very new field in research and students are often not coped with what needs to be done, so it does take a little longer for them to get them up to discipline. Recently we have a bunch of natural products, which we managed to isolate that are prominent for Africa and they're currently being used for SARS-CoV-2 treatment, which we're looking towards. To begin, go to the window icon on the monitor screen and click on it.

Select all apps and scroll down to locate the Schrodinger folder. Open the folder and click on the Maestro icon. Select open to launch the software.

To retrieve the protein structure, click on the file tab and select Get PDB from the popup menu. Enter the PDB code of choice in the text box and click the Download button. The selected PDB file will appear in the project window.

Alternatively, download the protein from the protein data bank by entering the PDB ID in the search box and clicking Download. In Maestro, navigate to the File tab and select Import Structures. In the import interface, locate the downloaded PDB file and click Import.

Now, select the protein structure and right click on it. Select the prepared protein, right click the mouse button, choose the Split option, and split into ligands, water, and others. Open the PubChem database and type the compound name in the search bar to download the chemical compound.

Review the available structures, click Download, and select 3D-Conformer to save the structure coordinates as a structured data file, or SDF. Click on the File tab in Schrodinger and select Import Structures. Navigate to the location where the SDF is saved and load the compound.

Click on Task in Schrodinger, type LigPrep in the search bar, and select it. Click Use Structures From to choose files from the workspace or project table. Select the preferred options in the LigPrep window and click Run to submit the job for ligand preparation.

Visualize the prepared ligands in the software window. Open the software for geometry optimization of the structures. Navigate to the File tab and select Open to choose the downloaded SDF from PubChem.

Navigate to the Calculate tab and select Gaussian Calculation Setup. In the job type tab, choose Optimization or Optimization plus Frequency. Now navigate to the Method tab and select the Quantum Chemistry method.

Choose the Cone Sham Global Hybrid Exchange Correlation Density Functional, Basis Set, Charge, and Spin of Choice from the dropdown menus. Go to the Title tab and enter a name for the compound under investigation. Navigate to the Link Zero tab and specify the memory limit and shared processors.

Untick the Full Path boxes. Click the Edit button at the bottom to save the Gaussian input file in the preferred location with a file name of choice as Gaussian Job File or GJF. Navigate to Tasks and select Receptor Grid Generation to detect the protein's active site bound to the core crystal ligand.

Click Pick to identify the ligand and then select the co-crystallized ligand. Click Run to submit the grid for generation. For molecular docking, go to Tasks, select Ligand Docking, and then choose Ligand Docking Glide Docking.

Then, load the grid file and select ligands from the workspace using the Use Ligand From option. Check the Display Receptor box at the top, add a suitable job name, and submit the job by clicking Run. From the Settings tab, choose the preferred docking precision method.

Configure constraints such as hydrogen bonds. After reviewing all settings, click Run to start the docking process. Review the docking results and compare the docking scores before and after optimizing the ligands.

Select a pair of the docked protein and the ligand complex from the workspace navigator. Navigate to Tasks, go to Ligand Designer, and click Analyze Workspace in the Ligand Designer window. To generate and evaluate new ligands, select Isostere Scanning from the workflow list, which implies the growing method that extends the ligand by adding fragments to existing molecular structures.

Analyze the docking results of the enumerated compounds and identify a compound with a more negative value than the co-crystallized compound minus 9.242. Next, click on the Task button and choose Desmond System Builder. In the System Builder panel, select the Solvation tab.

Choose the predefined solvent model that suits the protein ligand complex. Then select the box shape and the box size calculation method. Next, select the Ions tab and click recalculate to neutralize the system by adding counter ions and setting the desired solve concentration.

After the system preparation, view the project in the workspace. Select the protein ligand complex from the workspace navigator. Navigate to Task and choose Molecular Dynamics Desmond.

Load the ligand protein complex from the workspace in the Molecular Dynamics panel. Select the desired simulation timeline from the Simulation tab. Choose NPT as the ensemble class.

Name the job appropriately in the Molecular Dynamics panel. Write out the job and click on Close to exit the Molecular Dynamics window. Submit the written out job for molecular dynamics preparation via a local terminal.

Upon completion, open the completed job and continue the simulation time from the initial set timeline until the desired simulation time. For example, 100 nanoseconds or 200 nanoseconds. Open the Trajectory file and play the trajectory.

Visualize where the protein ligand complex is equilibrated and note the number of frames. Submit the job via Terminal. View the output file contents to analyze the generated results, and download the CSV file.

Open the CSV file and take note of the binding energy. Finally, use the shown equation and calculate the free binding energy of the complex by averaging the binding energy values determined for each snapshot within the MD simulation. A scatter plot shows the observed activity versus the predicted activity for class one of the QSAR model.

The graph represents the fitting between class one as the training set and the non nucleotide reverse transcriptase inhibitors as the test set to give a predictive activity value. The training set aligned well with the regression line while the test set had minor deviations. The forces of interactions between the protein in different ligands revealed hydrogen bonds with lysine 101 in all ligands.

Molecular dynamic simulations of the free protein stabilized after around 60 nanoseconds at an RMSD of around 3.5 Angstroms, confirming protocol reliability. Enumerated etravirine, which stabilized at 3.5. Angstroms, exhibited stronger and more stable binding at the active site of HIV one reverse transcriptase compared to etravirine which stabilized at 4.5 Angstroms.

The contact timeline of enumerated etravirine also indicated stronger and more stable interactions over time.