Learning Multimodal Graph-to-Graph Translation for Molecular Optimization

By Wengong Jin, Kevin Yang, Regina Barzilay, Tommi S. Jaakkola. 2018

267 citations recorded by the source corpus.

We view molecular optimization as a graph-to-graph translation problem. The goal is to learn to map from one molecular graph to another with better properties based on an available corpus of paired molecules. Since molecules can be optimized in different ways, there are multiple viable translations for each input graph. A key challenge is therefore to model diverse translation outputs. Our primary contributions include a junction tree encoder-decoder for learning diverse graph translations along with a novel adversarial training method for aligning distributions of molecules. Diverse output distributions in our model are explicitly realized by low-dimensional latent vectors that modulate the translation process. We evaluate our model on multiple molecular optimization tasks and show that our model outperforms previous state-of-the-art baselines.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Materials Chemistry, Generative Models, 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 Artificial Intelligence

In Computational Theory and Mathematics

In Materials Chemistry

In Generative Models

In Graph Learning

Citations in the corpus

Break-It-Fix-It: Unsupervised Learning for Program Repair (2021, background): “instances of this general setting include molecular design (Jin et al., 2019) which aims to improve the chemical properties (e.”
GraphQA: protein model quality assessment using graph convolutional networks (2020, background): “…et al., 2017; Li et al., 2018a), drug-drug interaction (Zitnik et al., 2018), drug-target interaction (Gao et al., 2018) molecular optimization (Jin et al., 2019), and generation of proteins, molecules and drugs (Ingraham et al., 2019a; You et al., 2018; Liu et al., 2018; Li et al., 2018b;…”
Therapeutics Data Commons: Machine Learning Datasets and Tasks for Drug Discovery and Development (2021, methodology): “Despite recent advances in generative modeling, existing methods in this area optimize ad-hoc heuristic oracles, such as QED (quantitative estimate of drug-likeness) and LogP (Octanolwater partition coefficient) [63, 169, 174].”
Deep Multimodal Fusion by Channel Exchanging (2020, background): “Joining the success of deep learning, multimodal fusion is recently specified as deep multimodal fusion by introducing end-to-end neural integration of multiple modalities [38], and it has exhibited remarkable benefits against the unimodal paradigm in semantic segmentation [29, 45], action recogniti”
Sample-Efficient Optimization in the Latent Space of Deep Generative Models via Weighted Retraining (2020, background): “Although many variants of these algorithms have been applied to real-world problems such as chemical design [29, 33, 39, 40, 8], the sample efficiency of this paradigm is currently unclear.”
Weisfeiler and Lehman Go Cellular: CW Networks (2021, background): “Junction Trees (JT), which conveniently represent singletons, bonds and rings as supernodes in a tree, have been used in molecular graph generation [43, 44].”
Differentiable Scaffolding Tree for Molecular Optimization (2021, unknown): “Typical algorithms include variational autoencoder (VAE), generative adversarial network (GAN), energy-based models, flow-based model [14, 21, 9, 37, 23, 17, 34, 38, 22, 49, 27, 5, 32, 1].”
Generalization and Representational Limits of Graph Neural Networks (2020, background): “Different types of updates are possible; we focus on a mean field update (Dai et al., 2016; Jin et al., 2018; 2019) h`v = φ ( W1xv +W2ρ( ∑ u∈N(v) g(h`−1u )) ) , (2) where φ and g are nonlinear transformations and xv ∈ Rr is the feature vector for v.”
Learning to simulate and design for structural engineering (2020, methodology): “…scientific domains including physics systems (Kipf et al., 2018; Sanchez-Gonzalez et al., 2018; Watters et al., 2017), fluid dynamics(Sanchez-Gonzalez et al., 2020), chemical molecules(Fout et al., 2017; Jin et al., 2018b; Do et al., 2019), and traffic networks (Guo et al., 2019; Cui et al., 2019).”
NAT: Neural Architecture Transformer for Accurate and Compact Architectures (2019, background): “Some researchers conduct architecture search by modeling the architecture as a graph [51, 20].”