Workflow Tutorials

This page is under construction.

Running an ESP workflow

The ESP workflow calculates the partial charges on atoms of a molecule. The charges are fit to the electrostatic potential at points selected according to the Merz-Singh-Kollman scheme, but other schemes supported by Gaussian can be used as well.

The ESP workflow performs the following steps:

%%{ init: { 'theme': 'base', 'themeVariables': { 'primaryTextColor': 'black', 'lineColor': 'lightgrey', 'secondaryColor': 'pink', 'tertiaryColor': 'lightgrey' } } }%% graph TD A[(Input Structure)] -->|Preprocessing| DFT DFT -->| | B[Geometry Optimization] B -->| | C[Frequency Calculation] C -->| | D[ESP Calculation] D -->|Postprocessing| E[(Output)] subgraph DFT B[Geometry Optimization] C[Frequency Calculation] D[ESP Calculation] end style A fill:#EBEBEB,stroke:#BB2528 style DFT fill:#DDEEFF,stroke:#DDEEFF,font-weight:bold style B fill:#fff,stroke-dasharray: 5, 5, stroke:#BB2528 style C fill:#fff,stroke-dasharray: 5, 5, stroke:#BB2528 style D fill:#fff,stroke:#BB2528 style E fill:#EBEBEB,stroke:#BB2528

Note

The geometry optimization and frequency calculation steps (marked with a dashed border in the above diagram) are optional. If the input structure is already optimized, the workflow will skip these steps.

In the following example, we will run the ESP workflow on a monoglyme molecule.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
from fireworks import LaunchPad

from mispr.gaussian.workflows.base.esp import get_esp_charges

lpad = LaunchPad.auto_load()

wf, _ = get_esp_charges(
    mol_operation_type="get_from_pubchem", # (1)!
    mol="monoglyme",
    format_chk=True,
    save_to_db=True,
    save_to_file=True,
    additional_prop_doc_fields={"name": "monoglyme"},
    tag="mispr_tutorial",
)
lpad.add_wf(wf) # (2)!

  1. mol_operation_type refers to the operation to be performed on the input to process the molecule.

    In this example, we are requesting to directly retrieve the molecule from PubChem by providing a common name for the molecule to be used as query criteria for searching the PubChem database via the mol input argument. For a list of supported mol_operation_type and the corresponding mol, refer to mispr.gaussian.utilities.mol.process_mol().

  2. Adds the workflow to the launchpad.

Download esp_tutorial.py.

Run the script using the following command:

python esp_tutorial.py

And then launch the job through the queueing system using the following command:

qlaunch rapidfire # (1)!

  1. This command can submit a large number of jobs at once or maintain a certain number of jobs in the queue.

The workflow will run and create a directory named C4H10O2 in the current working directory. The directory will contain the following subdirectories:

C4H10O2
├── Optimization
├── Frequency
├── ESP
├── analysis

Inside the Optimization, Frequency, and ESP subdirectories, you will find the Gaussian input and output files for the corresponding step. Inside the Optimization subdirectory, you will also find a “convergence.png” figure that shows the forces and displacement convergence during the course of the optimization.

../_images/convergence.png

The analysis subdirectory contains the results of the workflow in the form of a esp.json file. You can read the content of the esp.json file using the following commands:

1
2
3
4
5
6
import json

with open("C4H10O2/analysis/esp.json", "r") as f:
    esp = json.load(f)

print(esp["esp"])

This will output the partial charges on the atoms of the molecule:

{
"1": ["O", -0.374646],
"2": ["O", -0.373831],
"3": ["C", 0.132166],
"4": ["C", 0.132716],
"5": ["C", 0.034284],
"6": ["C", 0.031733],
"7": ["H", 0.033853],
"8": ["H", 0.034024],
"9": ["H", 0.034218],
"10": ["H", 0.034388],
"11": ["H", 0.070724],
"12": ["H", 0.03474],
"13": ["H", 0.03438],
"14": ["H", 0.034621],
"15": ["H", 0.071656],
"16": ["H", 0.034974],
}

Running a BDE workflow

Running an MD workflow

Running a hybrid workflow