GC-NORM

Description

The GC-NORM thermodynamic property prediction framework

Code content

• Norm descriptors calculation python script
• Molecular classification matlab script

Requirements

1. python 3.8
2. openbable
3. gaussian 16
4. matlab

Instructions for computing norm descriptors

The scripts come with two sample calculations for the molecules (CID 243 and 284). You can calculate the norm descriptors for any other target molecules according to the following instructions.

1. Add the open babel and gaussian 16 to the computer's environment variables so that it can be called from the command line to perform calculations.
2. Enter the CIDs of the target molecule into "cid.txt".
3. Go to the PubChem website (https://pubchem.ncbi.nlm.nih.gov/) and download the .sdf 3D geometry of the molecules corresponding to these CIDs.
4. Name the downloaded .sdf file as PubChem.sdf.
5. Open the powershell in the current directory and enter "python main.py" to automatically calculate the norm descriptor of the target molecules.
6. The final calculation results are saved in "descriptors.txt", and each row represents the 22 norm descriptors for a molecule.
7. Thermodynamic properties prediction based on norm descriptors using the parameters from "Norm_parameters" sheet of "Supplementary_data.xlsx".

Instructions for molecular classification

1. Enter the group segmentation results of the target molecules in "mol_group" sheet of the "mol_group.xlsx".
2. Use matlab to run "mol_type.m" in the current directory.
3. The classification results are stored in "result" variable in matlab, as shown in "result" sheet of the excel (You need to manually paste the classification results into excel to save them).