# GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation

@article{Shi2020GraphAFAF, title={GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation}, author={Chence Shi and Minkai Xu and Zhaocheng Zhu and Weinan Zhang and Ming Zhang and Jian Tang}, journal={ArXiv}, year={2020}, volume={abs/2001.09382} }

Molecular graph generation is a fundamental problem for drug discovery and has been attracting growing attention. The problem is challenging since it requires not only generating chemically valid molecular structures but also optimizing their chemical properties in the meantime. Inspired by the recent progress in deep generative models, in this paper we propose a flow-based autoregressive model for graph generation called GraphAF. GraphAF combines the advantages of both autoregressive and flow… Expand

#### Figures, Tables, and Topics from this paper

#### 73 Citations

GraphPiece: Efficiently Generating High-Quality Molecular Graph with Substructures

- Computer Science, Biology
- ArXiv
- 2021

This paper proposes a method to automatically discover common substructures, which are called graph pieces, from given molecular graphs by leveraging a variational autoencoder to generate molecules in two phases: piece-level graph generation followed by bond completion. Expand

MolGrow: A Graph Normalizing Flow for Hierarchical Molecular Generation

- Computer Science, Physics
- AAAI
- 2021

A new graph generative model—MolGrow, which is invertible and maps molecular structures onto a fixed-size hierarchical manifold, and proposes a fragment-oriented atom ordering that improves the model over commonly used breadth-first search ordering. Expand

MoFlow: An Invertible Flow Model for Generating Molecular Graphs

- Mathematics, Computer Science
- KDD
- 2020

The proposed MoFlow is a flow-based graph generative model to learn invertible mappings between molecular graphs and their latent representations and achieves state-of-the-art performance, which implies its potential efficiency and effectiveness to explore large chemical space for drug discovery. Expand

Molecular graph generation with Graph Neural Networks

- Mathematics, Computer Science
- ArXiv
- 2020

This paper introduces a sequential molecular graph generator based on a set of graph neural network modules, which it calls MG^2N^2, and indicates that the method outperforms very competitive baselines, and can be placed among the state of the art approaches for unconditional generation on QM9. Expand

Realistic molecule optimization on a learned graph manifold

- Computer Science, Physics
- ArXiv
- 2021

This work uses a hybrid approach, where the dataset distribution is learned using an autoregressive model while the score optimization is done using the Metropolis algorithm, biased toward the learned distribution, and shows that the resulting method, that is called learned realism sampling (LRS), produces empirically more realistic molecules and outperforms all recent baselines in the task of molecule optimization with similarity constraints. Expand

GraphDF: A Discrete Flow Model for Molecular Graph Generation

- Computer Science
- ICML
- 2021

This work proposes GraphDF, a novel discrete latent variable model for molecular graph generation based on normalizing flow methods that uses invertible modulo shift transforms to map discrete latent variables to graph nodes and edges. Expand

Graph Networks for Molecular Design

- 2020

Deep learning methods applied to chemistry can be used to accelerate the discovery of new molecules. This work introduces GraphINVENT, a platform developed for graph-based molecular design using… Expand

Graph networks for molecular design

- Computer Science, Physics
- Mach. Learn. Sci. Technol.
- 2021

This work compares six different GNN-based generative models in GraphINVENT, and shows that ultimately the gated-graph neural network performs best against the metrics considered here. Expand

Graph Neural Networks: A Review of Methods and Applications

- Computer Science, Mathematics
- AI Open
- 2020

A detailed review over existing graph neural network models is provided, systematically categorize the applications, and four open problems for future research are proposed. Expand

Flow-Based Models for Active Molecular Graph Generation

- 2020

We propose a framework using normalizing-flow based models, SELF-Referencing Embedded Strings, and active learning that generates a high percentage of novel, unique, and valid molecules and… Expand

#### References

SHOWING 1-10 OF 50 REFERENCES

MolecularRNN: Generating realistic molecular graphs with optimized properties

- Computer Science, Biology
- ArXiv
- 2019

MolecularRNN, the graph recurrent generative model for molecular structures, is presented, which generates diverse realistic molecular graphs after likelihood pretraining on a big database of molecules. Expand

Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation

- Computer Science, Mathematics
- NeurIPS
- 2018

Graph Convolutional Policy Network (GCPN) is proposed, a general graph convolutional network based model for goal-directed graph generation through reinforcement learning that can achieve 61% improvement on chemical property optimization over state-of-the-art baselines while resembling known molecules, and achieve 184% improved on the constrained property optimization task. Expand

Junction Tree Variational Autoencoder for Molecular Graph Generation

- Computer Science, Mathematics
- ICML
- 2018

The junction tree variational autoencoder generates molecular graphs in two phases, by first generating a tree-structured scaffold over chemical substructures, and then combining them into a molecule with a graph message passing network, which allows for incrementally expand molecules while maintaining chemical validity at every step. Expand

MolGAN: An implicit generative model for small molecular graphs

- Mathematics, Computer Science
- ArXiv
- 2018

MolGAN is introduced, an implicit, likelihood-free generative model for small molecular graphs that circumvents the need for expensive graph matching procedures or node ordering heuris-tics of previous likelihood-based methods. Expand

GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models

- Computer Science
- ICML
- 2018

The experiments show that GraphRNN significantly outperforms all baselines, learning to generate diverse graphs that match the structural characteristics of a target set, while also scaling to graphs 50 times larger than previous deep models. Expand

GraphNVP: an Invertible Flow-based Model for Generating Molecular Graphs

- Mathematics
- 2019

We propose GraphNVP, the first invertible, normalizing flow-based molecular graph generation model. We decompose the generation of a graph into two steps: generation of (i) an adjacency tensor and… Expand

GraphNVP: An Invertible Flow Model for Generating Molecular Graphs

- Mathematics, Computer Science
- ArXiv
- 2019

This work decomposes the generation of a graph into two steps: generation of an adjacency tensor and node attributes, which yields the exact likelihood maximization on graph-structured data, combined with two novel reversible flows. Expand

NeVAE: A Deep Generative Model for Molecular Graphs

- Computer Science, Physics
- AAAI
- 2019

A novel variational autoencoder for molecular graphs is proposed, whose encoder and decoder are specially designed to account for the above properties by means of several technical innovations. Expand

Neural Message Passing for Quantum Chemistry

- Computer Science
- ICML
- 2017

Using MPNNs, state of the art results on an important molecular property prediction benchmark are demonstrated and it is believed future work should focus on datasets with larger molecules or more accurate ground truth labels. Expand

Designing Random Graph Models Using Variational Autoencoders With Applications to Chemical Design

- Computer Science
- ArXiv
- 2018

Experiments reveal that the proposed variational autoencoder for graphs is able to learn and mimic the generative process of several well-known random graph models and can be used to create new molecules more effectively than several state of the art methods. Expand