# Exploring QSAR Models for Activity-Cliff Prediction

arXiv: https://arxiv.org/abs/2301.13644 · HTML: https://mlsanity.com/paper/2301.13644/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 86 references that are in the map.
- First posted on arXiv 31 Jan 2023.

## Builds on (most cited first, 16 of 16)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Graph Attention Networks](https://mlsanity.com/paper/1710.10903/) (2017)
- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [How Powerful are Graph Neural Networks?](https://mlsanity.com/paper/1810.00826/) (2018)
- [Neural Message Passing for Quantum Chemistry](https://mlsanity.com/paper/1704.01212/) (2017)
- [Optuna: A Next-generation Hyperparameter Optimization Framework](https://mlsanity.com/paper/1907.10902/) (2019)
- [Fast Graph Representation Learning with PyTorch Geometric](https://mlsanity.com/paper/1903.02428/) (2019)
- [Strategies for Pre-training Graph Neural Networks](https://mlsanity.com/paper/1905.12265/) (2019)
- [Convolutional Networks on Graphs for Learning Molecular Fingerprints](https://mlsanity.com/paper/1509.09292/) (2015)
- [ChemBERTa: Large-Scale Self-Supervised Pretraining for Molecular Property Prediction](https://mlsanity.com/paper/2010.09885/) (2020)
- [Molecular representation learning with language models and domain-relevant auxiliary tasks](https://mlsanity.com/paper/2011.13230/) (2020)
- [AttentionDDI: Siamese Attention-based Deep Learning method for drug-drug interaction predictions](https://mlsanity.com/paper/2012.13248/) (2020)
- [Using Molecular Embeddings in QSAR Modeling: Does it Make a Difference?](https://mlsanity.com/paper/2104.02604/) (2021)
- [Graph-augmented Convolutional Networks on Drug-Drug Interactions Prediction](https://mlsanity.com/paper/1912.03702/) (2019)
- [Beyond Textual Data: Predicting Drug-Drug Interactions from Molecular Structure Images using Siamese Neural Networks](https://mlsanity.com/paper/1911.06356/) (2019)

## Built on by (most cited first, 9 of 9)

- [Ranking over Regression for Bayesian Optimization and Molecule Selection](https://mlsanity.com/paper/2410.09290/) (2024)
- [Sort & Slice: A Simple and Superior Alternative to Hash-Based Folding for Extended-Connectivity Fingerprints](https://mlsanity.com/paper/2403.17954/) (2024)
- [AMPCliff: quantitative definition and benchmarking of activity cliffs in antimicrobial peptides](https://mlsanity.com/paper/2404.09738/) (2024)
- [FragNet: A Graph Neural Network for Molecular Property Prediction with Four Levels of Interpretability](https://mlsanity.com/paper/2410.12156/) (2024)
- [MaskMol: Knowledge-guided Molecular Image Pre-Training Framework for Activity Cliffs](https://mlsanity.com/paper/2409.12926/) (2024)
- [Mol-JEPA: A multimodal Joint Embedding Predictive Architecture for Molecules](https://mlsanity.com/paper/2608.22642/) (2026)
- [XIMP: Cross Graph Inter-Message Passing for Molecular Property Prediction](https://mlsanity.com/paper/2601.19037/) (2026)
- [The Geometry of Activity Cliffs: Representation Dependence and Multi-Scale Characterization of Activity Landscapes](https://mlsanity.com/paper/2605.30831/) (2026)
- [CliffRank: A Dual-Branch Framework for Activity-Cliff Ranking Prediction](https://mlsanity.com/paper/2609.01673/) (2026)

## Abstract (arXiv)

Pairs of similar compounds that only differ by a small structural modification but exhibit a large difference in their binding affinity for a given target are known as activity cliffs (ACs). It has been hypothesised that quantitative structure-activity relationship (QSAR) models struggle to predict ACs and that ACs thus form a major source of prediction error. However, a study to explore the AC-prediction power of modern QSAR methods and its relationship to general QSAR-prediction performance is lacking. We systematically construct nine distinct QSAR models by combining three molecular representation methods (extended-connectivity fingerprints, physicochemical-descriptor vectors and graph isomorphism networks) with three regression techniques (random forests, k-nearest neighbours and multilayer perceptrons); we then use each resulting model to classify pairs of similar compounds as ACs or non-ACs and to predict the activities of individual molecules in three case studies: dopamine receptor D2, factor Xa, and SARS-CoV-2 main protease. We observe low AC-sensitivity amongst the tested models when the activities of both compounds are unknown, but a substantial increase in AC-sensitivity when the actual activity of one of the compounds is given. Graph isomorphism features are found to be competitive with or superior to classical molecular representations for AC-classification and can thus be employed as baseline AC-prediction models or simple compound-optimisation tools. For general QSAR-prediction, however, extended-connectivity fingerprints still consistently deliver the best performance. Our results provide strong support for the hypothesis that indeed QSAR methods frequently fail to predict ACs. We propose twin-network training for deep learning models as a potential future pathway to increase AC-sensitivity and thus overall QSAR performance.