QSARtuna: QSAR model building with the optuna framework

Physicochemical properties, indices and descriptors for amino-acid sequences.

Exploring QSAR Models for Activity-Cliff Prediction

This workflow includes Quantitative Structure-Activity Relationship (QSAR) models to predict the hERG-related cardiotoxicity of chemicals.

Active learning workflow to train and fine-tune target molecular property predictors with chemist feedback in goal-oriented molecule generation.

Code for paper

Quantitative structure activity relationship models (QSAR models) using 6 molecular descriptors of 908 chemicals to predict quantitative acute aquatic toxicity (LC50 value) towards the fish Pimephales promelas (fathead minnow).

A modular inverse QSAR pipeline

Open Source, machine learning QSAR model with public data or your local data, The model utilises molecular descriptors as the independent variable, bioactivity as the dependent variable, random forest as a mathematical model.

Training data for "Prediction of clinically relevant drug-induced liver injury from structure using machine learning" (Hammann et al., J Appl Toxicol . 2019 Mar;39(3):412-419)

Research project predicting monomer pair reactivity ratios with QSAR models based on the U-V scheme and quantum chemical descriptors.

Estimate maximum performance bounds based on experimental errors for ML datasets

Exploring QSAR: From Data Collection to Structure-Activity Relationship Analysis

Code used in the elective course Advanced Computational Methods in Drug Discovery: AI and Physics-Based Simulations at Leiden University.

C. Tong List of Selected Publications & Abstracts

Developing a regression-based QSAR (quantitative structure-activity relationships) model to identify compounds that have 3-chymotrypsin-like protease (3CLpro) inhibitory activity.

A one stop destination of open source tools in Computer Aided Drug Design (CADD)

MCDCalc: Calculation of Markov Singular Values Molecular Descriptors Online Tool

This package facilitates developing Quantitative Structure-Activity Relationship (QSAR) models using the SEND database. It streamlines data acquisition, preprocessing, descriptor calculation, and model evaluation, enabling researchers to efficiently explore molecular descriptors and create robust predictive models.

Source code for the paper Cardoso-Silva, J., Papageorgiou, L. G. & Tsoka, S. (2019) Network-based piecewise linear regression for QSAR modelling. http://link.springer.com/10.1007/s10822-019-00228-6