1. Double-click the Materials Science icon
   • (No icon? See Starting Maestro)

Figure 2-1. Change Working Directory option.

2. Go to File > Change Working Directory
3. Find your directory, and click Choose
4. Pre-generated input and results files are included for running jobs or examining output. Download the zip file here: schrodinger.com/sites/default/files/s3/release/current/Tutorials/zip/autodpd.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 autodpd_tutorial, click Save
   • The project is now named autodpd_tutorial.prj

Figure 2-3. The entry list after importing.

We will construct a mixed polymer system containing polyethylene glycol (PEG) 50-mers and polylactic acid (PLA) 30-mers. These components are provided:

8. Go to File > Import Structures
9. Navigate to where you downloaded the provided tutorial files, choose AutoDPD_input_AA_polymers.maegz and click Open
   • A new entry group 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 two entries

If you are new to building all-atom polymers in Materials Science Maestro, visit the Building, Equilibrating and Analyzing Amorphous Polymers tutorial.

Note: The polymers were prepared with the Polymer Builder and are linear. Because the Tangled chain setting will be used in the subsequent build, it is no problem to construct the starting model in this way. In general, if you are using Snapped to grid or Amorphous states to build a disordered system, you should generate a molecule with a more reasonable conformation.

Figure 3-1. Selecting the entries and opening the 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 both PEG_50mer and PLA_30mer entries in the entry list
2. Go to Tasks > Materials > Classical Mechanics > Coarse Grain Models > Coarse-Grained Force Field Builder
   • The Coarse-Grained Force Field Builder panel opens
3. Click Load Selected Entries,
   • The two selected structures are loaded in the panel.

Figure 3-2. The Atomistic Input tab.

The Atomistic Input tab is used to define the all-atom system on which the coarse-grained system will be based.

4. For Number of molecules, input 64
   • The Components table should update automatically to indicate 32 of each polymer
5. In the Simulation section, set the Simulation time to 80 ns and the Trajectory recording interval to 200 ps
6. Go to the Map Atoms tab

Figure 3-3. The Map Atoms tab.

The Map Atoms tab is used to map atoms to particles, using SMARTS patterns.

7. For the first SMARTS group (1), input the following SMARTS pattern: [C-0X4][C-0X4][O-0X2][C-0X4][C-0X4][O-0X2][C-0X4][C-0X4][O-0X2][C-0X4][C-0X4][O-0X2][C-0X4][C-0X4][O-0X2]
   • This pattern represents 5 monomer units in a PEG chain
8. For the first SMARTS group (1), change the Particle name to EG
9. Change the Particle volume to 324.5 ų
   • The particle volume corresponds to the experimental PEG density of 1.127 g cm³ (see References)
10. Click Add SMARTS Group
    • A second SMARTS group (2) is added to the panel
11. For the second SMARTS group (2), input the following SMARTS pattern: [O-0X2][C-0X4]([C-0X4])[C-0X3](=[O-0X1])[O-0X2][C-0X4]([C-0X4])[C-0X3](=[O-0X1])[O-0X2][C-0X4]([C-0X4])[C-0X3](=[O-0X1])
    • This pattern represents 3 monomer units in a PLA chain
12. For the second SMARTS group (2), change the Particle name to LA
13. Change the Particle volume to 297.7 ų
    • The particle volume corresponds to the experimental PLA density of 1.206 g cm³ (see References)

Note: SMARTS patterns can also be generated directly from the workspace using the Use Workspace Selection button. If you are new to working with SMARTS patterns, see the Crosslinking Polymers tutorial which contains a detailed introduction.

Note: Use the Allow unique bonds and Allow atoms to belong to multiple particles for more flexible mapping, as described further in the help documentation

Note: The mapping scheme was adopted from the following study: Takhulee, A. et al. Journal of Polymer Research 2016, 24 (1), 8. In general, it is best to aim for similar sized particles such that particle volumes are within 50% of each other.

Figure 3-4. Viewing the CG mapping scheme

14. Click Show Mapped Molecules
    • This opens a new dialog panel while placing the CG structure in the workspace
15. Choose Mapping on all-atom system
    • The overlapping atoms in the CG structure are displayed in CPK style and the AA structure in wireframe.

This option lets us view the CG mapping scheme to the all-atom structure. We can change the color of the beads, and the structure will update accordingly. Upon closing the dialog, the loaded structure is removed from the workspace and the original structure reappears.

Figure 3-5. The FF Parameters tab.

16. Go to the FF Parameters tab

The FF (Force Field) Parameters tab is used to specify the parameters that are to be included in the coarse-grained simulation, provide initial values, and fix the values of parameters that do not need to be adjusted.

17. Click Populate Using Structure
    • Parameter types are added based on the all-atom disordered system structure
18. Check Use common mass and input 220
    • It is common in DPD to set a standard mass for all particles or subtypes of particles
19. Go to the Angle sub-tab

Figure 3-6. The FF Parameters tab (continued).

20. Click Ignore All Angles
    • The Angle parameters are removed from the table
21. Go to the Nonbonded sub-tab

Note: In this model, we will not parameterize angles. Typically, angles are beneficial for systems that have more fine-grained mappings, or systems composed of stiffer molecules.

Figure 3-7. The FF Parameters tab (continued 2).

22. Change the Cutoff distance to 9.91 Å
    • The cutoff distance is the primary length scale in a DPD simulation. The value chosen should be influenced by the reduced density and the particle volumes, which are displayed on the Map Atoms tab. The cutoff distance should be close to the value of [(reduced density)*(particle volume)]^⅓ for a chosen particle in the system. In this case the EG volume was used.
23. Go to the CG Simulation tab

Note: The reduced density is the number of particles within a volume unit of size (cutoff distance)³. The default value of the reduced density, 3.0, is sufficient for the system herein. Larger values of reduced density will increase the accuracy of a DPD model, in particular when fine-grained mapping schemes are used, but at the expense of increased computational intensity. Reduced values smaller than 3.0 are not recommended. Most studies employ values between 3.0 and 5.0.

Figure 3-8. The CG Simulation tab.

The CG (Coarse-Grained) Simulation tab is used to specify the conditions for the coarse-grained production simulations used for fitting the force-field parameters. For each iteration, the CG system is relaxed using a Brownie minimization step, followed by simulations of 100 ps at 10 K and 100 ps at 300 K in the NVT ensemble. Subsequently, the production simulation is performed in the NVT ensemble with the conditions specified in this tab:

24. Retain the defaults for the CG Simulation
25. Go to the Fitting tab

Figure 3-9. The Fitting tab and running the job.

The Fitting tab is used to specify the parameters for fitting the force field from the coarse-grained simulation results.

26. Change the Iterations to 20
    • Note in more complex systems, the fitting will benefit from more iterations
27. Change the Job name to cgff_builder_PEG_PLA
28. Adjust the job settings () as needed. This job requires a GPU host. The job can be completed in about 6 hours on a single GPU.
29. If you would prefer not to run the job, output files were provided and can be analyzed in Section 4. If you would like to run the job, click Run
30. Close the Coarse-Grained Force Field Builder panel

Figure 4-1. The output of the force field builder.

If you ran the job, the output will automatically incorporate, producing an entry titled amorphous_cg. Otherwise:

1. To import the results, go to File > Import Structures, navigate to the provided tutorial files and choose Section_04 > cgff_builder_PEG_PLA > cgff_builder_PEG_PLA-out.cms
   • An entry is added to the entry list titled amorphous_cg
   • The entry contains the final cell from the final coarse-grained simulation (the 20th iteration of the fitting procedure)

Figure 4-2. Builder Data tab

We can assess the quality of the fit using the Coarse Grain Forcefield Builder viewer panel.

2. Use the Workflow Action Menu (WAM) button () or go to Tasks > Materials > Classical Mechanics > Coarse Grain Models > Coarse-Grained Force Field Builder Results
   • The Coarse Grain Forcefield Builder Viewer panel panel opens
   • If the panel is not populated by default, click Load from Workspace

The Builder Data tab contains information about the input parameters.

Figure 4-3. Convergence tab

3. Go to the Convergence tab

The convergence tab shows how different parameters changed over the course of the fitting iterations.

4. Change Forcefield type to Bonds to view the bonding Req and k values as a function of iteration
   • We can see that the parameters are fairly well converged

Figure 4-4. Fit Quality tab: non-bonds, R-squared.

5. Go to the Fit Quality tab

The fit quality tab provides R-squared values for the quality of the fit as a function of iteration. It also contains a tab for viewing plots.

The non-bonds are displayed first. The data describe the pair distribution functions and their integrals for each non-bonded interaction in the CG system.

This data indicates how well the non-bonded parameters in the CG model reproduce the non-bonded interactions in the all-atom system. The coefficient of determination (R2), is close to 1 for each integral; however, these high R2 values do not necessarily imply a high quality of fit. The integrals (shown next) are the most important indicators.

Note: Changing the Iterations to display and Number of profiles to display options at the bottom of the panel can change the granularity of the data output.

Figure 4-5. Fit Quality tab: non-bonds, plot.

6. Go to the Plot tab

The plots contain pair distribution functions and their integrals for each interaction in the CG system: LA-LA, EG-EG and EG-LA. This data indicates how well the non-bonded parameters in the CG model reproduce the non-bonded interactions in the all-atom system

Good agreement can be observed visually between the all-atom (red line labeled Reference Analysis) and CG systems (lines from gray to black with the gradient representing iteration from first selected iteration to last selected). The comparisons of the integrals are shown, which are most important, since the integral is used directly in the fitting procedure.

In particular, the fit is based on the value of the integral at the cutoff distance (9.91 Å in this case), which can be observed to be similar for the all-atom and CG models.

Feel free to view the integral plots for the other Types (EG,LA and LA,LA) using the dropdown on the upper right. View the distribution function by changing to g(r) using the radio buttons on the upper left side of the panel.

Figure 4-6. Fit Quality tab: Bonds, R-squared.

7. Return to the R-squared tab and change the Forcefield type to Bonds
   • Again, R-squared values are provided now for the bonded parameters
8. Set Number of profiles to display to 5

Figure 4-7. Fit Quality tab: Bonds, plot.

9. Go to the Plot tab

This data indicates how well the bonded parameters in the CG model reproduce the bonded interactions in the all-atom simulation.

Good agreement can be observed visually between the all-atom (red line labeled Reference Analysis) and CG systems (lines from gray to black with the gradient representing iteration from first selected iteration to last selected).

Figure 4-8. Fit Quality tab: Bonds, plot; LA-LA.

10. In the Type dropdown menu, choose LA,LA

We can see the successful fitting of the LA,LA bonded parameter as well.

Feel free to explore the various plots available for assessing the quality of the fitting. 

Figure 4-9. Inspecting the .json file.

The cgff_builder_PEG_PLA_cgff-out.json file contains the final force field parameters from the automated procedure. It can be accessed from the output directory.

The output directory also contains a file mapped_cg_structures.maegz with individual input structures for the separate system components. These can be used to construct new systems of various concentrations.