Convolutional Networks on Graphs for Learning Molecular Fingerprints
By David Kristjanson Duvenaud, Dougal Maclaurin, Jorge Aguilera-Iparraguirre, Rafael Gómez‐Bombarelli, Timothy D. Hirzel, Alán Aspuru-Guzik, Ryan P. Adams. 2015
3,666 citations recorded by the source corpus.
We introduce a convolutional neural network that operates directly on graphs. These networks allow end-to-end learning of prediction pipelines whose inputs are graphs of arbitrary size and shape. The architecture we present generalizes standard molecular feature extraction methods based on circular fingerprints. We show that these data-driven features are more interpretable, and have better predictive performance on a variety of tasks.
Subfields: Computational Theory and Mathematics, Computer Vision and Pattern Recognition, Materials Chemistry, Molecular Biology, Graph Learning
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computational Theory and Mathematics
- Tommi S. Jaakkola
- Regina Barzilay
- Jure Leskovec
- William L. Hamilton
- Gisbert Schneider
- Wengong Jin
- Kyle Swanson
- Max Welling
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Ming-Hsuan Yang
- Trevor Darrell
- Xiaogang Wang
- Jitendra Malik
- Andrea Vedaldi
- Andrew Zisserman
In Materials Chemistry
- Tommi S. Jaakkola
- Regina Barzilay
- William L. Hamilton
- Jure Leskovec
- Shitong Luo
- Jian Tang
- Wengong Jin
- Max Welling
In Molecular Biology
- William L. Hamilton
- Jure Leskovec
- Tommi S. Jaakkola
- Regina Barzilay
- Rex Ying
- Stephan Günnemann
- Yang Zhang
- Jian Tang
In Graph Learning
- William L. Hamilton
- Jure Leskovec
- Stephan Günnemann
- Matthias Fey
- Rex Ying
- Michael M. Bronstein
- Jian Tang
- Marinka Zitnik
Citations in the corpus
- Semi-Supervised Classification with Graph Convolutional Networks (2016, methodology): “Firstly, we introduce a simple and well-behaved layer-wise propagation rule for neural network models which operate directly on graphs and show how it can be motivated from a first-order approximation of spectral graph convolutions (Hammond et al., 2011).”
- Graph Attention Networks (2017, methodology): “In some cases, this requires learning a specific weight matrix for each node degree (Duvenaud et al., 2015), using the powers of a transition matrix to define the neighborhood while learning weights for each input channel and neighborhood degree (Atwood & Towsley, 2016), or extracting and normalizing…”
- Inductive Representation Learning on Large Graphs (2017, neutral): “We then describe how the GraphSAGE model parameters can be learned using standard stochastic gradient descent and backpropagation techniques (Section 3.2).”
- Gated Graph Sequence Neural Networks (2015, methodology): “Duvenaud et al. (2015) also consider convolutional like operations on graphs, building a learnable, differentiable variant of a successful graph feature.”
- VAIN: Attentional Multi-agent Predictive Modeling (2017, neutral): “[17] for fingerprinting of chemical molecules.”
- Discriminative Embeddings of Latent Variable Models for Structured Data (2016, neutral): “Duvenaud et al. (2015) used CNN to learn the circulant fingerprints for graphs from end to end.”
- Knowledge Graph Convolutional Networks for Recommender Systems (2019, background): “To handle the neighborhoods with varying size and maintain the weight sharing property of CNN, researchers propose learning a weight matrix for each node degree [6], extracting locally connected regions from graphs [13], or sampling a fixed-size set of neighbors as the support size [7].”
- Neural Message Passing for Quantum Chemistry (2017, methodology): “While neural networks have been applied in a variety of situations (Merkwirth & Lengauer, 2005; Micheli, 2009; Lusci et al., 2013; Duvenaud et al., 2015), they have yet to become widely adopted.”
- Directional Message Passing for Molecular Graphs (2020, methodology): “Recently, these methods have largely been superseded by graph neural networks, which do not require any hand-crafted features but learn representations solely based on the atom types and coordinates molecules (Duvenaud et al., 2015; Gilmer et al., 2017; Schütt et al., 2017; Unke & Meuwly, 2019).”
- Graph Convolutional Neural Networks for Web-Scale Recommender Systems (2018, methodology): “Following on this work, a number of authors proposed improvements, extensions, and approximations of these spectral convolutions [6, 10, 11, 13, 18, 21, 24, 29, 31], leading to new state-of-the-art results on benchmarks such as node classification, link prediction, as well as recommender system task”
- Relational inductive biases, deep learning, and graph networks (2018, methodology): “For illustration, consider the task of predicting the center Duvenaud et al., 2015).”
- Flexible Neural Representation for Physics Prediction (2018, positive): “Our work draws inspiration from and extends that of Chang et al. [11] and Battaglia et al. [4], which in turn use ideas from graph-based neural networks [39, 44, 9, 30, 22, 14, 13, 24, 8, 40].”
- The More You Know: Using Knowledge Graphs for Image Classification (2016, background): “There has been a lot of work in end-to-end learning on graphs or neural network trained on graphs [30, 3, 6, 11, 24, 22, 9, 21].”
- Molecular graph convolutions: moving beyond fingerprints (2016, methodology): “Recently, Duvenaud et al. (2015) presented an architecture trying to accomplish many of the same goals as this work.”
- Unsupervised Inductive Whole-Graph Embedding by Preserving Graph Proximity (2019, methodology): “Across the years, many methods have been proposed for the generation of graph-level embeddings from node-level embeddings, with the most popular ones including: • Sum/Average: The graph embedding is simply the summation or average of the node embeddings (Duvenaud et al., 2015).”