OR

1. Double click the Maestro or Materials Science icon to start Maestro or MS Maestro
   • No icon? See Starting Maestro
   • This tutorial uses MS Maestro, but this workflow can be performed in Maestro or MS Maestro. Use whichever interface you are comfortable with or typically use for your projects.

Figure 2-1. Change Working Directory option.

2. Go to File > Change Working Directory
3. Find your directory, and click Choose
4. Pre-generated files are included for running jobs or examining output. Download the zip file here: schrodinger.com/sites/default/files/s3/release/current/Tutorials/zip/automartini_protein.zip
5. 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. Save Project panel.

6. Go to File > Save Project As
7. Change the File name to automartini_protein_tutorial, click Save
   • The project is now named automartini_protein_tutorial.prj

Figure 2-3. Importing the input structure.

8. Go to File > Import Structures
9. Navigate to where you downloaded the provided tutorial files (presumably in your working directory), choose multistage_simulation_trp_cage_excipients > multistage_simulation_trp_cage_excipients-out.cms
10. Click Open
    • A new entry is added to the entry lista simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion containing the entry disordered_system_trp_cage_excipients_all_components_amorphous system
    • The imported structure is the equilibrated MD system from the Simulating Complex Protein Solutions tutorial and it contains the 2JOF protein, Tween 20, sucrose, water, chlorine anions, and sodium cations

Figure 3-1. Opening the Coarse-Grained Mapping panel.

1. In the entry lista simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion, 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 the only entry
2. Go to Tasks > Materials > Classical Mechanics> Coarse Grain Models > Coarse-Grained Mapping
   • The Coarse-Grained Mapping panel opens

Figure 3-2. Running the Coarse-Grained Mapping calculation.

3. Check Use automated CG mapping
4. Select Martini
5. Change the Job name to coarse_grained_mapping_trp_cage_excipients
6. Adjust the job settings as needed. This job requires a CPU host. The job can be completed in about 5 minutes.
   • If you receive a warning that says “Nucleic acids are not fully supported for automated Martini mapping, we recommend contacting Schrödinger Support,” click OK  

Figure 3-3. The mapped particles.

The panel is populated with the mapped particles and the workspace is updated when the job completes.

7. Go to the Restraints tab  

Figure 3-4. Viewing the protein’s restraints.

8. Check Unique Molecules
   • Checking unique molecules allows one to examine how individual types of molecules are mapped as opposed to seeing the mapping of the whole system
9. Select all the rows and view the restraints in the workspace
   • Restraints can be added or removed and then saved for future use

Figure 3-5. Exporting the mapped structures to the CGFF Builder.

10. Go back to the Particle Types tab
11. Check Systems
12. Export mapped structures to CGFF Builder
    • When either Systems or Unique Molecules is selected the atomistic and CG versions can be saved to the project table. By hitting Export mapped structures to Project table these can be manually transferred to the CGFF builder and the molecules can also be used to build systems (e.g., with the disordered system builder).
    • When Systems is selected the mapping information can be transferred directly to the CGFF builder panel. This will automatically open up the CGFF builder.  

Figure 3-6. The CGFF Builder panel.

The Coarse-Grained Force Field Builder panel automatically opens and the disordered_system_trp_cage_excipients_all_components entry is automatically loaded. This entry contains both the CG system and its associated all-atom trajectory.

13. Select Martini for the Coarse-graining type
14. Check Use force field and select Martini_solution from the dropdown menu
15. Go to the Mapped Atoms tab  

Figure 3-7. Automatic mapping of CG particles

This tab is autopopulated with the information transferred from the Coarse-Grained Mapping panel. Each particle contains a complex SMARTS string.

16. Go to the FF Parameters tab

 

Figure 3-8. Populating the CG particle types.

17. Click Populate using structures
    • The Particle subtab contains the list of CG bead names and the Martini bead type
    • This can take a minute or two
18. Go to the Nonbonded subtab
    • The non-bonded interaction pairs are listed in the table
    • The interaction strength (ε) is a fitting parameter while the particle size will be kept constant by default.
19. Click Import from Force Field
    • Encrypted parameter values are listed for fixed protein-protein interactions, see the help documentation for more information

Figure 3-9. Setting up the CG simulation parameters.

20. Go to the CG Simulation tab
    • For each iteration a 20 ns NPT CG MD simulation will be performed for the fitting procedure. We will keep the default settings.

Figure 3-10. Setting and running the job.

21. Go to the Fitting tab
22. Analyze last 40% of simulation
    • For each iteration the last 40% of the trajectory will be used for the fitting procedure
23. Change the Job name to cgff_builder_trp_cage_excipients
24. Adjust the job settings as needed. This job requires a GPU host. The job can be completed in about 10 hours.
25. If you would like to run the job yourself, click Run. Otherwise, go to File > Import Structures, navigate to the provided tutorial files and OpenSection_03 > cgff_builder_trp_cage_excipients > cgff_builder_trp_cage_excipients-out.cms
26. Close the Coarse-Grained Force Field Builder panel

Figure 3-11. Output structure.

Once the job is completed or after importing, 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 cgff_builder_trp_cage_excipients-out.cms

Figure 4-1. Using the WAM button to open the results panel.

1. 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 cgff_builder_trp_cage_excipients-out entry in the entry lista simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion
2. Use the Workflow Action Menu (WAM) button () to open the Coarse-Grained Force Field Results panel
   • Alternatively go to Tasks > Materials > Classical Mechanics> Coarse Grain Models > Coarse-Grained Force Field Builder Results
   • The Coarse-Grained Force Field Builder Viewer panel opens

 

Figure 4-2. Summary of the particle types.

In the Builder Data tab is a summary of input parameters set in the parameters from the simulation. Specifically, the Particle name, SMARTS string for each bead type, Charge and Mass of the CG particles. Note that the WF type particle appears as one of the CG particle types in the system.

3. Go to the Convergence tab

Figure 4-3. Convergence of Nonbonded forcefield parameters.

This plot shows the force field parameters fitted using the Coarse-Grained Force Field Builder Panel, plotted against the number of fitting iterations to illustrate how they evolve during the fitting process. By default, the non-bonded interactions are displayed. The Epsilon values have a clear trend for the first 15-20 iterations. The values become more stable (graph plateauing) as more iterations are accumulated. Using a longer CG simulation time (60 ns rather than 20ns) might help get more stable values.

 

The initial sigma values are recommended since they were not included in this fitting process.

 

Later in this section, we will look at ASP_S1 (ASP sidechain) with E2 (PEG group from tween 20). There are two ASP groups and 80 E2 groups in the system so there may be moderate sampling for this interaction.

 

Figure 4-4. Convergence of ASP_S1,E2 Nonbonded forcefield parameters.

4. Select ASP_S1,E2 for the Type
   • E2 is a moderately small site with a short bond-length to adjacent sites.
   • In Martini 2, the standard rules tend to overestimate the strength of non-bonded interactions involving particles with short bonds.
   • We see that the Epsilon drops from the initial value during the iterations resulting in weaker interactions.

Figure 4-5. Convergence of the bond parameters.

5. Change the Forcefield type to Bonds
   • In the final iterations, the parameters should consistently oscillate around a stable value

Figure 4-6. Convergence of bond angle parameters.

6. Change the Forcefield type to Angles
7. Select E2(C2),(C3,1,1)C2(O4),(C1)E2 for the Type
   • Here, we are focusing on an E2-C2-E2 angle potential (specifically E2(C2),(C3,1,1)C2(O4),(C1)E2 ) from the Tween 20 molecule
   • There are four Tween 20 molecules and each has only one of this particular type of E2-C2-E2 angle 
   • This provides relatively little averaging during sampling leading to larger noise
8. Go to the Fit Quality tab

Figure 4-7. Fitting score for the automated CG mapping.

The R-squared value for the fitting parameters are displayed compared to the all-atom reference at specific fitting iterations. You can explore the fitting score for other interaction types from the dropdown options.

9. Change the Number of profiles to display to 3
10. Go to the Plot subtab

 

Note: R-squared parameters serve as a guide for users to determine whether further inspection of the plots is warranted. While high R-squared values generally indicate good fits, low R-squared values can also represent acceptable fits in some cases depending on the total fit. Additionally, small molecules grouped together to form a particle are often difficult to fit accurately, making it preferable to use a fixed standard value in such instances.

Figure. 4-8. Comparing the nonbond distribution with the reference system.

This tab visualizes how the fitting parameters in the CG model compare against the all-atom reference.

11. Select ASP_S1,E2 for the Type
12. Select g(r)
    • This specifies the pair distribution function for the parameter detailed in the Type option menu
13. Set the Number of profiles to display to 5
    • Initially, there were excessive contacts between ASP_S1 and E2.
    • The reference g(r) is small and noisy, suggesting this interaction is uncommon, potentially unimportant, and undersampled in the simulation.
    • Longer CG simulations could reduce noise in the fitting process, but this may not be essential due to the small reference g(r), indicating this interaction might not be critical.

Figure 4-9. Comparing the angle distribution with the reference system.

14. Change the Forcefield type to Angles
    • The distribution of the bond angles for the reference and the CG simulations are plotted
15. Select E2(C2),(C3,1,1)C2(O4),(C1)E2 for the Type
16. Set the Number of profiles to display to 3
    • The plot displays a bimodal reference distribution, which is colored red.
    • The CG model, which employs a harmonic angle potential, typically exhibits a single peak.
    • The fitting process should produce a coarse-grained (CG) distribution that closely matches the average and standard deviation of the reference distribution. Specifically, the average of the CG distribution should be similar to the reference average, and its standard deviation should also approximate that of the reference.
    • Each of the four Tween 20 molecules contains only one instance of this specific E2-C2-E2 angle.
    • This results in minimal averaging during sampling, which leads to increased noise, as evident in the distributions.  

Feel free to explore other options in the panel for further analysis. You can save the CG forcefield data to the Schrödinger directory. This can then be used to run further simulations with different compositions and system sizes of the same components.