Skip to content

A machine learning QSPR model predicting the octanol-water partition coefficient (logP) using PaDEL physicochemical descriptors

Notifications You must be signed in to change notification settings

luiraym/RAYLOGP

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

56 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RAYLOGP

A machine learning QSPR model predicting the octanol-water partition coefficient (logP) using PaDEL physicochemical descriptors.

This repository contains the experimental files used to compare different methods of representing a molecule in-silico to effectively model a chemical space to logP as described by Lui et al. (2020) JCAMD 34: 523-534.

The overall best model, 'RAYLOGP', was a stochastic gradient descent-optimised linear regression with 1,438 physicochemical descriptors calculated in PaDEL. External validation was performed by predicting for 11 protein kinase inhibitor fragment-like molecules prepared for the 2019 SAMPL6 LogP Prediction Challenge, returning an average RMSE of 0.49 log units (submission ID: 'hdpuj').

About

A machine learning QSPR model predicting the octanol-water partition coefficient (logP) using PaDEL physicochemical descriptors

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages