MSCSol 1.0.5

A tool for predicting the solubility of small molecule drugs.

Using MSCSol to predict moleculer solubility

MSCSol

Our trained model has been uploaded to PyPI, accessible through this link (https://pypi.org/project/MSCSol/). We've included detailed installation instructions and usage guidelines, making it easy to obtain prediction results by inputting SMILES strings.

Installation

pip install MSCSol==1.0.5

Package needed

see https://github.com/ZiyuFanCSU/MSCSol and you need to download the required data files for the program and place them in the directory where you are running MSCSol, including trained.pt and fingerprint.csv.

Quick Start

import MSCSol

MSCSol.pred(<your_SMILES>)

Note

It will take some time to calculate the molecular signatures, so please be patient for a while. Also note that dipole moment features are not used here as they cannot be obtained directly by calling the code.

The training data was restricted to molecular weights less than or equal to 504, LogS values greater than or equal to -8, and experimental temperatures of 20-25 degrees Celsius, so if the molecule does not apply to the above conditions, the prediction results may have a large deviation.

In addition, due to the computational requirements of the node vector feature of the GVP-GNN, the input molecule atom number must be greater than or equal to 3.

In addition, four temporary files will be generated at runtime, named img_MSCSol.png, shadow_MSCSol_1.png, shadow_MSCSol_2.png and shadow_MSCSol_3.png. Please make sure not to have the same file name as yours to avoid accidental deletion.

Update log

1.0.1--1.0.5 Modifying some bugs.

1.0.0 First release.

Contact

We thank all the researchers who contributed to this work.

If you have any questions, please contact [e-mail].