Abstract:Molecular featurisation refers to the transformation of molecular data into numerical feature vectors. It is one of the key research areas in molecular machine learning and computational drug discovery. Recently, message-passing graph neural networks (GNNs) have emerged as a novel method to learn differentiable features directly from molecular graphs. While such techniques hold great promise, further investigations are needed to clarify if and when they indeed manage to definitively outcompete classical molecular featurisations such as extended-connectivity fingerprints (ECFPs) and physicochemical-descriptor vectors (PDVs). We systematically explore and further develop classical and graph-based molecular featurisation methods for two important tasks: molecular property prediction, in particular, quantitative structure-activity relationship (QSAR) prediction, and the largely unexplored challenge of activity-cliff (AC) prediction. We first give a technical description and critical analysis of PDVs, ECFPs and message-passing GNNs, with a focus on graph isomorphism networks (GINs). We then conduct a rigorous computational study to compare the performance of PDVs, ECFPs and GINs for QSAR and AC-prediction. Following this, we mathematically describe and computationally evaluate a novel twin neural network model for AC-prediction. We further introduce an operation called substructure pooling for the vectorisation of structural fingerprints as a natural counterpart to graph pooling in GNN architectures. We go on to propose Sort & Slice, a simple substructure-pooling technique for ECFPs that robustly outperforms hash-based folding at molecular property prediction. Finally, we outline two ideas for future research: (i) a graph-based self-supervised learning strategy to make classical molecular featurisations trainable, and (ii) trainable substructure-pooling via differentiable self-attention.

Expressivity and Generalization: Fragment-Biases for Molecular GNNs

Towards Bridging Generalization and Expressivity of Graph Neural Networks

Breaking the Expression Bottleneck of Graph Neural Networks

Breaking the Expressive Bottlenecks of Graph Neural Networks

Is fragment-based graph a better graph-based molecular representation for drug design? A comparison study of graph-based models

Will More Expressive Graph Neural Networks do Better on Generative Tasks?

Beyond weisfeiler-lehman: A quantitative framework for GNN expressiveness

Fragment-based Pretraining and Finetuning on Molecular Graphs

Utilizing Edge Features in Graph Neural Networks Via Variational Information Maximization

On the Expressive Power of Geometric Graph Neural Networks

Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting

FragNet: A Graph Neural Network for Molecular Property Prediction with Four Layers of Interpretability

A Complete Expressiveness Hierarchy for Subgraph GNNs via Subgraph Weisfeiler-Lehman Tests

Investigating Graph Neural Networks and Classical Feature-Extraction Techniques in Activity-Cliff and Molecular Property Prediction

Graph Neural Network Expressivity and Meta-Learning for Molecular Property Regression

HiGNN: Hierarchical Informative Graph Neural Networks for Molecular Property Prediction Equipped with Feature-Wise Attention

Rethinking the expressive power of gnns via graph biconnectivity

On the Scalability of GNNs for Molecular Graphs

Gini in a Bottleneck: Sparse Molecular Representations for Graph Convolutional Neural Networks

An Empirical Study of Realized GNN Expressiveness