# Topological regression as an interpretable and efficient tool for quantitative structure-activity relationship modeling

## TL;DR

This paper introduces topological regression, a new QSAR method that is as accurate as deep learning but more interpretable for drug discovery.

## Contribution

Topological regression offers a novel, interpretable, and efficient QSAR modeling approach with competitive predictive performance.

## Key findings

- Topological regression achieves equal or better performance than deep learning QSAR models on 530 ChEMBL datasets.
- The method provides intuitive interpretation by mapping chemical space to activity space through approximate isometry.
- TR is computationally efficient and statistically grounded, making it suitable for molecular design tasks.

## Abstract

Quantitative structure-activity relationship (QSAR) modeling is a powerful tool for drug discovery, yet the lack of interpretability of commonly used QSAR models hinders their application in molecular design. We propose a similarity-based regression framework, topological regression (TR), that offers a statistically grounded, computationally fast, and interpretable technique to predict drug responses. We compare the predictive performance of TR on 530 ChEMBL human target activity datasets against the predictive performance of deep-learning-based QSAR models. Our results suggest that our sparse TR model can achieve equal, if not better, performance than the deep learning-based QSAR models and provide better intuitive interpretation by extracting an approximate isometry between the chemical space of the drugs and their activity space.

Quantitative structure-activity relationships (QSAR) models are widely used in drug discovery, but have limitations in their interpretability and accuracy near activity cliffs. Here the authors use a topological regression framework to increase QSAR interpretability and efficiency.