TUHH-TVT / openCOSMO-RS_py

GNU General Public License v2.0
56 stars 11 forks source link

Questions about COSMO-RS Implementation in opencosmorspy #9

Open VN-jsjeon opened 2 weeks ago

VN-jsjeon commented 2 weeks ago

I have two questions regarding the COSMO-RS implementation using openCOSMO-RS_py:

1. Reference State Verification

I'm using the following code to calculate COSMO-RS descriptors:

import numpy as np
from opencosmorspy import Parameterization, COSMORS

crs = COSMORS(par='default_orca')
crs.par.calculate_contact_statistics_molecule_properties = True
solute = ['/home/jeon/test/cosmors-10.solute.orcacosmo']
crs.add_molecule(solute)
x = np.array([1.0])
T = 298.15  # K
crs.add_job(x=x, T=T, refst='cosmo')
results = crs.calculate()
solvation_energies_kcal = {}
for key in results['enth'].keys():
    if key.startswith('aim_A'):
        solvation_energies_kcal[key] = results['enth'][key] / 4184  # J/mol to kcal/mol

print("\nSolvation free energies (kcal/mol):", solvation_energies_kcal)
print("\nDetailed solvation energies (kcal/mol):")
print(f"Misfit contribution (aim_A_mf): {solvation_energies_kcal['aim_A_mf'][0][0]:.2f}")
print(f"H-bond contribution (aim_A_hb): {solvation_energies_kcal['aim_A_hb'][0][0]:.2f}")
print(f"Total solvation energy (aim_A_int): {solvation_energies_kcal['aim_A_int'][0][0]:.2f}")

Question: 1-1. I'm using a .orcacosmo file generated from ORCA6 calculations - is this the correct input file format for calculating COSMO-RS descriptors using opencosmorspy? 1-2. I'm using refst='cosmo' - can you confirm if this is correct for calculating solvation free energies?

2. COSMO-RS Descriptors for Conformer Clustering

I'm trying to extract specific COSMO-RS descriptors for use in conformer clustering analysis. Specifically, I'm interested in:

These descriptors would be used to cluster and analyze different conformers based on their surface properties.

Question: 2-1. Does opencosmorspy support extracting these descriptors for conformer analysis? If so, how can I access them? 2-2. If not, are there alternative methods to obtain these values for conformer clustering?

Any guidance would be greatly appreciated!

simonmb commented 2 weeks ago

1-1. I'm using a .orcacosmo file generated from ORCA6 calculations - is this the correct input file format for calculating COSMO-RS descriptors using opencosmorspy? Yes 1-2. I'm using refst='cosmo' - can you confirm if this is correct for calculating solvation free energies? Yes

2-1. Does opencosmorspy support extracting these descriptors for conformer analysis? If so, how can I access them? Yes, You don't even need opencosmorspy for this. The SigmaProfileParser does this already. So you can either take it from here or from the conformer pipeline repo. It has everything you need. 2-2. If not, are there alternative methods to obtain these values for conformer clustering? It is available.

What do you mean by conformer clustering? So finding conformers that are similar to bin them? That should be doable.

Also another information: Unfortunately the development of the python version has not kept up with all of the changes of the cpp version. I was not the original developer of the python version and all efforts are put into the cpp version. If anyhow possible I would recommend using that version instead. I know it is a bit more difficult to get working, but I would think it's worth it.

VN-jsjeon commented 2 weeks ago

@simonmb Thank you so much! However, I'am a bit confused and would like to ask some additional questions.

After running OpenCOSMO calculations with ORCA6, I obtained several output files (listed below):

cosmors-9.bibtex              cosmors-9.solute_cpcm.cpcm_corr      cosmors-9.solute_cpcm.property.txt  cosmors-9.solute_vac.inp           cosmors-9.solvent_cpcm.densities
cosmors-9.json                cosmors-9.solute_cpcm.densities      cosmors-9.solute.orcacosmo          cosmors-9.solute_vac.lastout       cosmors-9.solvent_cpcm.densitiesinfo
cosmors-9_out.json            cosmors-9.solute_cpcm.densitiesinfo  cosmors-9.solute_vac.bibtex         cosmors-9.solute_vac.property.txt  cosmors-9.solvent_cpcm.gbw
cosmors-9.property.txt        cosmors-9.solute_cpcm.gbw            cosmors-9.solute_vac.densities      cosmors-9.solvent_cpcm.bibtex      cosmors-9.solvent_cpcm.inp
cosmors-9.solute_cpcm.bibtex  cosmors-9.solute_cpcm.inp            cosmors-9.solute_vac.densitiesinfo  cosmors-9.solvent_cpcm.cpcm        cosmors-9.solvent_cpcm.lastout
cosmors-9.solute_cpcm.cpcm    cosmors-9.solute_cpcm.lastout        cosmors-9.solute_vac.gbw            cosmors-9.solvent_cpcm.cpcm_corr   cosmors-9.solvent_cpcm.property.txt

To compute the COSMO-RS descriptors, I’d like to confirm which files are required as input for the COSMO-RS descriptor calculations.

Specifically, should I use the cosmors-9.solute.orcacosmo file only, or do I need to include cosmors-9.solvent.orcacosmo as well?

Currently, I am using chloroform as the solvent.

Thank you for your assistance!

simonmb commented 2 weeks ago

To calculate the sigma profile and to be able to calculate e. g. sigma moments you can use the SigmaProfileParser on the orcscosmo file for which you want to calculate the descriptors. If you need orcacosmo files for files not directly available ftok ORCA and specifically if you need several conformers for one structure you need to modify our conformer pipeline in the other repository.

Did this answer your question?