Exploring Protein Binding Sites with Mixed-Solvent Molecular Dynamics

Tutorial Created with Software Release: 2025-2
Topics: Biologics Drug Discovery, Small Molecule Drug Discovery, Structure Prediction & Target Enablement
Products Used: Desmond

Tutorial files

11 MB
This tutorial is written for use with a 3-button mouse with a scroll wheel.
Words found in the Glossary of Terms are shown like this: Workspacethe 3D display area in the center of the main window, where molecular structures are displayed

 

Tip: You can hover over a glossary term to display its definition. You can click on an image to expand it in the page.
Abstract:

 

In this tutorial, you will learn how to set up and run a series of MD simulations on a protein with different water cosolvent mixtures in a single job using the mixed solvent MD method to identify and characterize probable binding sites for small organic molecules on a protein surface. You will analyze the character of the predicted binding sites and compare their position to that of a known ligand. Lastly, you will compare the predicted binding site with results from SiteMap.

 

Tutorial Content

  1. Introduction to Mixed Solvent MD Simulations

  1. Creating Projects and Importing Structures

  1. Setting Up and Running the MxMD Simulation

  1. Analyzing the Results

  1. Conclusion and References

  1. Glossary of Terms

1. Introduction to Mixed Solvent MD Simulations

Identification of potential ligand binding sites on a drug target is a critical step in structure-based drug design. Protein binding sites can generally be characterized by binding hotspots on the target surface that have a high propensity for ligand binding and typically consist at least in part of solvent-exposed, hydrophobic amino acid residues. This specific arrangement allows organic molecules with partially hydrophobic properties to effectively compete with the bulk solvent for the binding hotspots through a combination of enthalpic and entropic contributions, where loosely bound water molecules on the hydrophobic protein surface can be displaced with minimal energy penalty.

Mixed solvent molecular dynamics (MxMD) is a computational method used to identify and characterize likely binding sites for small organic molecules on surfaces and interfaces. The solute can be a protein, DNA or RNA. The technique involves performing unbiased all-atom molecular dynamics (MD) simulations in binary explicit solvent mixtures, typically water and small organic and water miscible probes (cosolvents).

Figure 1: Schematic setup of a cosolvated system.

These cosolvents mimic the binding of ligand counterparts to the solute, allowing exploration of potential interaction sites. Through its underlying dynamic approach, MxMD takes into account the full flexibility of solute, solvent and cosolvent, and thus the competition between solvent and cosolvent probes. By analyzing probe occupancy, binding hotspots can be mapped, cryptic pockets can be revealed, and the "druggability" of a target surface can be assessed.

Figure 2: Schematic representation of cosolvent probes binding at a hotspot.

In this tutorial, you will use MxMD to find and assess potential ligand binding sites in the human E3 ubiquitin-protein ligase Mdm2. The structure used was co-crystalized with an inhibitor. You will learn how to set up and run an MxMD job from the Maestro graphical user interface (GUI), which uses Schrödinger's MD engine Desmond with the underlying force field OPLS4. This tutorial will guide you through the analysis of the results, comparing the position and character of a predicted binding site with a known ligand, and a comparison of MxMD and SiteMap results.

2. Creating Projects and Importing Structures

At the start of the session, change the file path to your chosen Working Directorythe location where files are saved in Maestro to make file navigation easier. Each session in Maestro begins with a default Scratch Projecta temporary project in which work is not saved, closing a scratch project removes all current work and begins a new scratch project, which is not saved. A Maestro project stores all your data and has a .prj extension. A project may contain numerous entries corresponding to imported structures, as well as the output of modeling-related tasks. Once a project is created, the project is automatically saved each time a change is made.

Structures can be imported from the PDB, or from your Working Directorythe location where files are saved using File > Import Structures, and are automatically added to the Entry Lista simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion and Project Tabledisplays the contents of a project and is also an interface for performing operations on selected entries, viewing properties, and organizing structures and data. The Entry Lista simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion is located to the left of the Workspacethe 3D display area in the center of the main window, where molecular structures are displayed. The Project Tabledisplays the contents of a project and is also an interface for performing operations on selected entries, viewing properties, and organizing structures and data can be accessed by Ctrl+T (Cmd+T) or Window > Project Table if you would like to see an expanded view of your project data.

  1. Double-click the Maestro icon.

Figure 2-1. Changing the Working Directory via the option in the main window.

  1. Go to File > Change Working Directory.
    • The Change Directory panel opens.
  2. Browse to the directory you want to use as your Working Directorythe location where files are saved, and click Select Folder.
  3. Pre-generated input and results files are included for running jobs or examining output. Download the zip file here: https://www.schrodinger.com/sites/default/files/s3/release/current/Tutorials/zip/mxmd.zip.
  4. After downloading the zip file, unzip the contents in your Working Directorythe location where files are saved for ease of access throughout the tutorial.

Figure 2-2. Saving the project via the Save Project As dialog box.

  1. Go to File > Save Project As.
  2. Change the File name to MxMD_4JV7, and change the Location to your Working Directorythe location where files are saved, then click Save.
    • The project is now named MxMD_4JV7.prj and is saved in your Working Directorythe location where files are saved.

Figure 2-3. Importing the prepared structure of the protein.

  1. Go to File > Import Structures….
    • The Import dialog box opens.
  2. Navigate to your Working Directory, select the file 4JV7_prepared.mae and select Open. Confirm by clicking Import.
    • A banner appears and a group is added to your Entry Lista simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion.

Feel free to familiarize yourself with the protein before moving on, e.g. by choosing a visualization style that is most helpful to you.

Structure files obtained from the PDB, vendors, and other sources often lack necessary information for performing modeling-related tasks. Typically, these files are missing hydrogens, have incorrect or missing bond order assignments, charge states, side chain orientation, or are missing loop regions. In order to make these structures suitable for modeling tasks, the Protein Preparation Workflow is used to resolve common structural issues.

In this tutorial, the system (Chain A of the PDB entry 4JV7) has already been prepared. If you are following along with your own structure, make sure it is fully prepared before processing to the setup of the MxMD simulation. The Introduction to Structure Preparation and Visualization tutorial as well as the Best Practices for Protein Preparation can guide you through the basics of the process. For more information tailored to MD applications, see the Introduction to All-Atom Molecular Dynamics Simulations with Desmond tutorial. For MxMD it is essential to build missing loops, otherwise artificial hotspots may be detected at the place of the missing loop (see MxMD Best Practices).

3. Setting Up and Running the MxMD Simulation

In this section, you will learn how to set up and run an MxMD job with the Schrödinger suite via the Maestro graphical user interface with the Mixed Solvent MD panel, starting with a prepared protein system. Although MxMD jobs are based on MD simulations, you do not need to set up a Desmond model system yourself or worry about minimizing and equilibrating the system. These steps are performed automatically in the MxMD workflow.

The input structure is solvated with a 7Å layer of the cosolvent, which in turn is solvated in water so that the volume ratio of the cosolvent to the whole volume of the binary mixture is approximately 5%. This system setup is performed individually for each selected cosolvent probe times the number of simulations per probe, which gives the number of individual simulations for the job. Every simulation setup is performed with different cosolvent configurations. To learn more about setting up and running simple MD simulations with Desmond, see the Introduction to All-Atom Molecular Dynamics Simulations with Desmond tutorial. After the setup, every system is equilibrated in several stages, followed by an additional MD stage for production, from which frames are saved for analysis.

As an alternative to the GUI, MxMD jobs can be run from the command line, which provides more options for customizing your jobs. For more information on usage and options, see the MxMD command help or the Running MxMD Simulations from the Command Line.

The MxMD job requires a prepared structure without any ligands, solvent and other molecules to sample the solutes surface with the cosolvent probes. You may choose to retain essential metal ions and water molecules involved in binding or other biological function.

Figure 3-1. Splitting the prepared protein entry into ligand, water and protein.

  1. Right-click on the entry 4JV7 - prepared, go to Split > Into Ligands, Water, Other.
    • The original entry is split into individual entries for the ligand, waters, and the protein.
  2. Select(1) the atoms are chosen in the Workspace. These atoms are referred to as "the selection" or "the atom selection". Workspace operations are performed on the selected atoms. (2) The entry is chosen in the Entry List (and Project Table) and the row for the entry is highlighted. Project operations are performed on all selected entries and includethe entry is represented in the Workspace, the circle in the In column is blue the entry 4JV7 - prepared_protein.

Figure 3-2. The Mixed Solvent MD panel.

  1. Go to Tasks > Browse > Desmond > Mixed Solvent MD….
    • The Mixed Solvent MD panel opens.
  2. Make sure that for Task, the Setup option is selected.

Figure 3-3. Loading the solute’s structure and job settings.

  1. For Solute from, choose Workspace (included entry) from the dropdown menu, then click Load.
    • The solute’s structure is loaded into the panel.

The default cosolvent probes are acetonitrile, isopropanol, and pyrimidine. They work well for most protein solutes by mapping out regions that interact well with polar/amphipathic, H-bond/small hydrophobic, and six-membered aromatic moieties, respectively. There are 28 additional cosolvent probes readily available in the suite, which can be chosen from the dropdown menu in the panel. Note that these probes have been selected and optimized for the detection of hotspots in protein solutes. For nucleotide solutes or more tailored applications, you may wish to use your own custom probes, which can be added via a recommended workflow (command line only).

Other available settings are the number of simulations per cosolvent probe, and the equilibration and productive simulation time. The default values are estimates, which work well for most systems, but you may want to increase them for more thorough equilibration and sampling in the case of complex solutes, e.g. those undergoing large-scale dynamics. You can find the details of the equilibration and simulation protocol here. If your system is a membrane protein, a lipid bilayer of your choice can be automatically placed around your protein via the build membrane option, provided that its orientation is OPM-compatible.

Figure 3-4. The Mixed Solvent MD - Job Settings dialog box.

  1. Click the cog.
    • The Mixed Solvent MD - Job Settings dialog box opens.
  2. Change the job name to mixed_solvent_4JV7.
  3. Optional: Select a suitable CPU Host and a Linux-based GPU subhost.
  4. Click OK.
  5. Optional: Click Run to start the job.

The MxMD job is now ready to run. Because the workflow utilizes Desmond MD simulations, the corresponding subjobs can only run on Linux-based hosts with GPUs (see Desmond System Requirements).Pre-generated result files of the system can be found in the zip file you downloaded at the beginning. Feel free to run the job if you have access to suitable hardware.

4. Analyzing the Results

In this section, you will analyze the results of an MxMD job by visually inspecting the predicted binding hotspots and comparing their position to the position of a known ligand, as well as looking at some hotspot properties. In the last step, you will compare the results of MxMD with pregenerated SiteMap results to compare both methods.

For each probe, frames from the productive MD simulations are extracted and aligned with the starting structure. The probe occupancy or probe density is calculated on a grid for all extracted frames and can be visualized in a contoured plot similar to an electron density. These probe occupancy maps are then clustered with a 3Å cutoff into so-called spots. Overlapping spots from different probes define a hotspot.

The output of an MxMD job is automatically saved in the form of a compressed Maestro project archive (.prjzip). Additional output like MD frames can generally be found in a results folder, which is not included in the tutorial zip file here due to the size of the files.

4.1 Analyzing the MxMD Results

Please note that calculations may have been performed with an earlier version of the software and the results may not be exactly the same as those you produced in this tutorial. Especially for methods with a statistical component, such as MD simulations, an exact reproduction of a trajectory and its properties is only possible if the same random seed, software and hardware configuration are used. With sufficient sampling, all properties of the ensemble should be reproducible, while the properties of individual frames and the exact course of a trajectory may differ.

Figure 4-1. Loading the MxMD results into the panel.

  1. Switch the Task in the top of the MxMD panel to the View result option.
    • The panel changes from the setup mode to the analysis mode.
  2. For Import results from choose File.
  3. Click Browse… and choose the file mixed_solvent_4JV7.prjzip.
    • The results are loaded into the panel and the Workspacethe 3D display area in the center of the main window, where molecular structures are displayed, a new entry group is added to your Entry Lista simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion.

Figure 4-2. Occupancy map of the cosolvent probes on the protein surface.

  1. In addition to the mixed_solvent_4JV7_input entry, fix the ligand entry (4JV7 - prepared_ligand) in the Workspacethe 3D display area in the center of the main window, where molecular structures are displayed by double-clicking the includethe entry is represented in the Workspace, the circle in the In column is blue circle.
  2. Choose a visualization that is most helpful to you. For example, we have displayed the protein as grey ribbons, the ligand in a green ball-and-stick representation, and added a mostly transparent surface to the protein.

The current settings show the occupation of all probes on the surface of the protein: Acetonitrile in orange, isopropanol in blue, and pyrimidine in yellow. You can turn them off and on individually via the Display Settings in the MxMD panel or via the Manage Surfaces panel.

Figure 4-3. The Solvent hotspots table in the MxMD panel.

  1. In the MxMD panel, click the Surface option for hotspot 0.
    • Hotspot 0 is shown as a grey surface.
  2. Click the Probes option for hotspot 0, then go to Display Options and for Molecules select only acetonitrile.
    • The acetonitrile molecules from all frames for hotspot 0 are shown in the Workspacethe 3D display area in the center of the main window, where molecular structures are displayed.
  3. Repeat the last step for the other two probes.
  4. Investigate the other hotspots.

The solvent hotspots table contains information for all detected hotspots: The hotspot index, number of probes at the hotspot (minimum 2), MxMD score, hotspot surface area and volume, and probe types at the hotspot. The MxMD score is a measure of occupancy for all probes in a hotspot and can be used to identify regions of high probe density, which are likely to be of biological interest. Consider all sites with MxMD scores above 10000 as promising candidates, regardless of their rank-order.

Find more information on how hotspots are obtained and the MxMD score is calculated here.

 

Figure 4-4. Hotspots 0 (left) and 2 (right) on the protein surface.

The first hotspot (index 0) aligns well with the ligand. It contains all three cosolvent probes occupied by the ligand and is by far the largest hotspot in terms of MxMD score, surface area, and volume. Three subpockets are detected and found to be big enough to accommodate 6-membered aromatic rings (pyrimidine probe, yellow). The two subpockets occupied by the bromine substituted phenyl rings can be characterized as amphipathic (acetonitrile probe, orange) with a potential for hydrogen bonding (isopropanol probe, blue). Additionally, hotspot 0 and hotspot 2 show occupancy near one of the phenyl rings and the central ring of the ligand, respectively. This space is not occupied by the ligand and hints at potential growth space at these positions. All other hotspots are significantly smaller in terms of surface area and volume, and show smaller MxMD scores. This makes them unlikely binding sites. 

4.2 Comparing MxMD and SiteMap Results

SiteMap is another tool for identifying and characterizing potential binding sites. It uses a static approach on a single protein structure. SiteMap has already been run on the same system and identified only a single potential binding site.

Figure 4-5. Overlaying the results of MxMD and SiteMap for the ligand binding site.

  1. Go to File > Import Structures…, select the file sitemap_4JV7_out.maegz, click Open and then confirm the import.
    • The SiteMap results are loaded. Note that the color coding of the MxMD results is automatically adjusted.
  2. Includethe entry is represented in the Workspace, the circle in the In column is blue the following entries in the Workspacethe 3D display area in the center of the main window, where molecular structures are displayed: 4JV7 - prepared_ligand; mixed_solvent_4JV7_input; sitemap_4JV7_site_1.
  3. Compare the identified binding sites.

Both SiteMap and MxMD detect the ligand binding site, although MxMD shows a more complete coverage of the subpockets in terms of occupied volume as well as more detailed characterization of the subpockets in terms of interactions. SiteMap only partially overlaps or misses some subpockets. This increased level of detail from MxMD simulations presents a clear benefit of its dynamic approach over a static approach such as SiteMap.

Learn more about SiteMap for identifying and characterizing potential binding sites in the Target Analysis with SiteMap and WaterMap tutorial.

5. Conclusion and References

In this tutorial, you learned how to set up and run an MxMD job from the Maestro GUI to identify and characterize probable binding sites for small organic molecules on a protein surface. You analyzed the positions and character of predicted cosolvent hotspots and compared them to a known ligand in the structure. Lastly, you compared the predicted binding site by MxMD with results from SiteMap. MxMD successfully detected and assessed the ligand binding site with all of its subpockets, while the SiteMap results only partially resemble the subpockets.

6. Glossary of Terms

Entry List - a simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion

Included - the entry is represented in the Workspace, the circle in the In column is blue

Project Table - displays the contents of a project and is also an interface for performing operations on selected entries, viewing properties, and organizing structures and data

Recent actions - This is a list of your recent actions, which you can use to reopen a panel, displayed below the Browse row. (Right-click to delete.)

Scratch Project - a temporary project in which work is not saved, closing a scratch project removes all current work and begins a new scratch project

Selected - (1) the atoms are chosen in the Workspace. These atoms are referred to as "the selection" or "the atom selection". Workspace operations are performed on the selected atoms. (2) The entry is chosen in the Entry List (and Project Table) and the row for the entry is highlighted. Project operations are performed on all selected entries

Working Directory - the location where files are saved

Workspace - the 3D display area in the center of the main window, where molecular structures are displayed