Psikit: a thin wrapper library for Psi4 and RDKit

Inspired from the entry:Calculate HOMO and LUMO with Psi4

Install RDKit and Psi4 from Conda

conda install -c psi4 psi4
conda install -c rdkit rdkit
conda install -c conda-forge debtcollector

Install resp from github repository (resp from conda doesn't work)

git clone https://github.com/cdsgroup/resp.git
cd resp
pip install .

Install Psikit

Psikit is under development but you can install the current version of Psikit from pypi or conda.

via conda

conda install -c iwatobipen psikit

via pip

pip install psikit

via pip from github

pip install git+https://github.com/Mishima-syk/psikit

Testing Psikit

pytest --disable-warnings -v

Usage

Single point calcuration

from psikit import Psikit
pk = Psikit()
pk.read_from_smiles("c1ccccc1")
print("SCF Energy: ", pk.energy())
print("HOMO: ", pk.HOMO)
print("LUMO: ", pk.LUMO)
x, y, z, total = pk.dipolemoment
print("SCF Total Dipole Moment: {}".format(total))
# SCF Energy:  -230.712279648862
# HOMO:  -0.32848562009092513
# LUMO:  0.1456515222506689
# SCF Total Dipole Moment: 3.292464934070545e-05

Structure optimization

pk = Psikit()
pk.read_from_smiles("c1ccccc1")
print("Optimized SCF Energy: ", pk.optimize())
# Optimizer: Optimization complete!
# Optimized SCF Energy:  -230.71352354223438

Calculate RESP Charge

# REF http://ambermd.org/tutorials/advanced/tutorial1/files/resp_paper_jacs.pdf
pk = Psikit()
pk.read_from_smiles("CC(=O)O")
pk.optimize()
# Optimizer: Optimization complete!
# -227.82180859253418
pk.calc_resp_charges()
# array([-0.32506898,  0.83672649, -0.61924915, -0.66135715,  0.10450057,
#    0.10478188,  0.10780051,  0.45186584])

for atom in pk.mol.GetAtoms(): 
    print(atom.GetSymbol(), "ESP:{}\tRESP:{}".format(atom.GetProp("EP"), atom.GetProp("RESP"))) 

# C ESP:-0.49662019588648315	RESP:-0.3250689814483399
# C ESP:0.91473263536048643		RESP:0.83672648554100837
# O ESP:-0.63823808477114718	RESP:-0.61924915363703359
# O ESP:-0.6763331997116846		RESP:-0.66135714989354499
# H ESP:0.14625849864628995		RESP:0.10450056830656008
# H ESP:0.14578513969681847		RESP:0.10478187811883517
# H ESP:0.1530843954112609		RESP:0.1078005104750676
# H ESP:0.45133081125445906		RESP:0.45186584253744722

### Compute Mulliken charges and Lowdin charges

pk = Psikit()
pk.read_from_smiles("CC(=O)O")
pk.optimize() # or pk.energy()

pk.calc_mulliken_charges()
# array([-0.42203029,  0.72794785, -0.55419051, -0.59333358,  0.16369722,
#    0.1636994 ,  0.15462075,  0.35958916])

pk.calc_lowdin_charges()
#array([-0.30006577,  0.33900448, -0.35983788, -0.28463832,  0.12439944,
#    0.12810672,  0.11935266,  0.23367866])

Rendering Molecular Orbitals

from psikit import Psikit
pk = Psikit()
pk.read_from_smiles("c1ccccc1")
pk.optimize(basis_sets="scf/sto-3g")
pk.view_on_pymol() # launch pymol as a RPC server in advance, just type "pymol -R"

Adding RDKit mol object to Psikit object directly

from psikit import Psikit
pk = Psikit()
pk.mol = your_mol_object

Jupyter notebook

• RESP charge
• Torsion Scan
• Rendering Orbital with VMD
• Rendering Orbital in PyMol
• Charge Comparison
• SAPT
• FSAPT

License

Code released under the BSD license.