{"type":"ScholarlyArticle","title":"MolHF: A Hierarchical Normalizing Flow for Molecular Graph Generation","authors":[{"name":"Yiheng Zhu","slug":"yiheng-zhu","given":"Yiheng","family":"Zhu","affiliation":"College of Computer Science and Technology , Zhejiang University"},{"name":"Zhenqiu Ouyang","slug":"zhenqiu-ouyang","given":"Zhenqiu","family":"Ouyang","affiliation":"Polytechnic Institute , Zhejiang University"},{"name":"Ben Liao","slug":"ben-liao","given":"Ben","family":"Liao","affiliation":"Tencent Quantum Laboratory , Tencent"},{"name":"Jialu Wu","slug":"jialu-wu","given":"Jialu","family":"Wu","affiliation":"College of Pharmaceutical Sciences , Zhejiang University"},{"name":"Yixuan Wu","slug":"yixuan-wu","given":"Yixuan","family":"Wu","affiliation":"School of Public Health , Zhejiang University"},{"name":"Chang-Yu Hsieh","slug":"chang-yu-hsieh","given":"Chang-Yu","family":"Hsieh","affiliation":"College of Pharmaceutical Sciences , Zhejiang University"},{"name":"Tingjun Hou","slug":"tingjun-hou","given":"Tingjun","family":"Hou","affiliation":"College of Pharmaceutical Sciences , Zhejiang University"},{"name":"Jian Wu","slug":"jian-wu","given":"Jian","family":"Wu","affiliation":"School of Public Health , Zhejiang University; Second Affiliated Hospital School of Medicine , Zhejiang University {zhuyiheng2020 , oyzq ,"}],"abstract":"Molecular de novo design is a critical yet challenging task in scientific fields, aiming to design novel molecular structures with desired property profiles. Significant progress has been made by resorting to generative models for graphs. However, limited attention is paid to hierarchical generative models, which can exploit the inherent hierarchical structure (with rich semantic information) of the molecular graphs and generate complex molecules of larger size that we shall demonstrate to be difficult for most existing models. The primary challenge to hierarchical generation is the non-differentiable issue caused by the generation of intermediate discrete coarsened graph structures. To sidestep this issue, we cast the tricky hierarchical generation problem over discrete spaces as the reverse process of hierarchical representation learning and propose MolHF, a new hierarchical flow-based model that generates molecular graphs in a coarse-to-fine manner. Specifically, MolHF first generates bonds through a multi-scale architecture, then generates atoms based on the coarsened graph structure at each scale. We demonstrate that MolHF achieves state-of-the-art performance in random generation and property optimization, implying its high capacity to model data distribution. Furthermore, MolHF is the first flow-based model that can be applied to model larger molecules (polymer) with more than 100 heavy atoms. The code and models are available at https://github.com/violet-sto/MolHF.","doi":"10.24963/ijcai.2023/556","url":"https://ijcai.org/proceedings/2023/556","pdf":"https://ijcai.org/proceedings/2023/0556.pdf","markdown":"https://ijcai.org/proceedings/2023/556.md","bibtex":"https://ijcai.org/proceedings/2023/bibtex/556","volume":{"slug":"2023","title":"Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2023","url":"https://ijcai.org/proceedings/2023/","year":2023,"location":"Macao, SAR","dates":"19-25 August 2023","editors":"Edith Elkind","isbn":"978-1-956792-03-4","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2023"},"section":"Multidisciplinary Topics and Applications","track":"Main Track","number":556,"pages":{"first":5002,"last":5010},"keywords":["Machine Learning: ML: Probabilistic machine learning","Machine Learning: ML: Sequence and graph learning","Multidisciplinary Topics and Applications: MDA: Health and medicine"],"videos":[],"references":[{"position":1,"raw":"[Behrmann et al., 2021] Jens Behrmann, Paul Vicol, Kuan- Chieh Wang, Roger Grosse, and Jörn-Henrik Jacobsen. Understanding and mitigating exploding inverses in invert- ible neural networks. In AISTATS, 2021.","title":"Understanding and mitigating exploding inverses in invertible neural networks","authors":"Jens Behrmann, Paul Vicol, Kuan-Chieh Wang, Roger Grosse, Jörn-Henrik Jacobsen","year":"2021","venue":"AISTATS","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[De Cao and Kipf, 2018] Nicola De Cao and Thomas Kipf. MolGAN: An implicit generative model for small molec- ular graphs. ICML workshop, 2018.","title":"MolGAN: An implicit generative model for small molecular graphs","authors":"Nicola De, Cao, Thomas Kipf","year":"2018","venue":"ICML workshop","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Degen et al., 2008] Jörg Degen, Christof Wegscheid- Gerlach, Andrea Zaliani, and Matthias Rarey. On the art of compiling and using'drug-like'chemical fragment spaces. ChemMedChem: Chemistry Enabling Drug Discovery, 3(10):1503-1507, 2008.","title":"On the art of compiling and using'drug-like'chemical fragment spaces","authors":"Degen","year":"2008","venue":"ChemMedChem: Chemistry Enabling Drug Discovery","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Dinh et al., 2014] Laurent Dinh, David Krueger, and Yoshua Bengio. Nice: Non-linear independent compo- nents estimation. arXiv preprint arXiv:1410.8516, 2014.","title":"Nice: Non-linear independent components estimation","authors":"Laurent Dinh, David Krueger, Yoshua Bengio","year":"2014","venue":null,"doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Dinh et al., 2016] Laurent Dinh, Jascha Sohl-Dickstein, and Samy Bengio. Density estimation using real nvp. arXiv preprint arXiv:1605.08803, 2016.","title":"Density estimation using real nvp","authors":"Laurent Dinh, Jascha Sohl-Dickstein, Samy Bengio","year":"2016","venue":null,"doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Gilmer et al., 2017] Justin Gilmer, Samuel S Schoenholz, Patrick F Riley, Oriol Vinyals, and George E Dahl. Neural message passing for quantum chemistry. In ICML, pages 1263-1272. PMLR, 2017.","title":"Neural message passing for quantum chemistry","authors":"Justin Gilmer, S Samuel, Patrick F Schoenholz, Oriol Riley, George E Vinyals, Dahl","year":"2017","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Gómez-Bombarelli et al., 2018] Rafael Gómez- Bombarelli, Jennifer N Wei, David Duvenaud, José Miguel Hernández-Lobato, Benjamín Sánchez- Lengeling, Dennis Sheberla, Jorge Aguilera-Iparraguirre, Timothy D Hirzel, Ryan P Adams, and Alán Aspuru- Guzik. Automatic chemical design using a data-driven continuous representation of molecules. ACS central science, 4(2):268-276, 2018.","title":"Automatic chemical design using a data-driven continuous representation of molecules","authors":"Rafael Gómez-Bombarelli, Jennifer N Wei, David Duvenaud, José Miguel Hernández-Lobato, Benjamín Sánchez-Lengeling, Dennis Sheberla, Jorge Aguilera-Iparraguirre, Timothy D Hirzel, Ryan P Adams, Alán Aspuru-Guzik","year":"2018","venue":"ACS central science","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Goodfellow et al., 2014] Ian Goodfellow, Jean Pouget- Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Gen- erative adversarial nets. Advances in neural information processing systems, 27, 2014.","title":"Generative adversarial nets","authors":"Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, Yoshua Bengio","year":"2014","venue":"Advances in neural information processing systems","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Ho et al., 2019] Jonathan Ho, Xi Chen, Aravind Srinivas, Yan Duan, and Pieter Abbeel. Flow++: Improving flow- based generative models with variational dequantization and architecture design. In ICML, 2019.","title":"Flow++: Improving flowbased generative models with variational dequantization and architecture design","authors":"Jonathan Ho, Xi Chen, Aravind Srinivas, Yan Duan, Pieter Abbeel","year":"2019","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Huang et al., 2019] Zilong Huang, Xinggang Wang, Lichao Huang, Chang Huang, Yunchao Wei, and Wenyu Liu. Cc- net: Criss-cross attention for semantic segmentation. In ICCV, 2019.","title":"Ccnet: Criss-cross attention for semantic segmentation","authors":"Zilong Huang, Xinggang Wang, Lichao Huang, Chang Huang, Yunchao Wei, Wenyu Liu","year":"2019","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Irwin et al., 2012] John J Irwin, Teague Sterling, Michael M Mysinger, Erin S Bolstad, and Ryan G Coleman. Zinc: a free tool to discover chemistry for biology. JCIM, 52(7):1757-1768, 2012.","title":"Zinc: a free tool to discover chemistry for biology","authors":"Teague John J Irwin, Sterling, M Michael, Erin S Mysinger, Ryan G Bolstad, Coleman","year":"2012","venue":"JCIM","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Jin et al., 2018] Wengong Jin, Regina Barzilay, and Tommi Jaakkola. Junction tree variational autoencoder for molec- ular graph generation. In ICML, 2018.","title":"Junction tree variational autoencoder for molecular graph generation","authors":"Jin","year":"2018","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Jin et al., 2020] Wengong Jin, Regina Barzilay, and Tommi Jaakkola. Hierarchical generation of molecular graphs us- ing structural motifs. In ICML, 2020.","title":"Hierarchical generation of molecular graphs using structural motifs","authors":"Jin","year":"2020","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Karras et al., 2018] Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. Progressive growing of gans for im- proved quality, stability, and variation. In ICLR, 2018.","title":"Progressive growing of gans for improved quality, stability, and variation","authors":"Tero Karras, Timo Aila, Samuli Laine, Jaakko Lehtinen","year":"2018","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Kingma and Dhariwal, 2018] Diederik P Kingma and Pra- fulla Dhariwal. Glow: generative flow with invertible 1× 1 convolutions. In NeurIPS, 2018.","title":"Glow: generative flow with invertible 1× 1 convolutions","authors":"P Diederik, Prafulla Kingma, Dhariwal","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Kingma and Welling, 2013] Diederik P Kingma and Max Welling. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013.","title":"Auto-encoding variational bayes","authors":"P Diederik, Max Kingma, Welling","year":"2013","venue":null,"doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Kipf and Welling, 2017] Thomas N. Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. In ICLR, 2017.","title":"Semi-supervised classification with graph convolutional networks","authors":"Thomas N Kipf, Max Welling","year":"2017","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Kong et al., 2022] Xiangzhe Kong, Wenbing Huang, Zhix- ing Tan, and Yang Liu. Molecule generation by principal subgraph mining and assembling. In Alice H. Oh, Alekh Agarwal, Danielle Belgrave, and Kyunghyun Cho, editors, NeurIPS, 2022.","title":"Molecule generation by principal subgraph mining and assembling","authors":"Kong","year":"2022","venue":null,"doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Kuznetsov and Polykovskiy, 2021] Maksim Kuznetsov and Daniil Polykovskiy. Molgrow: A graph normalizing flow for hierarchical molecular generation. In AAAI, 2021.","title":"Molgrow: A graph normalizing flow for hierarchical molecular generation","authors":"Polykovskiy Kuznetsov, Maksim Kuznetsov, Daniil Polykovskiy","year":"2021","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Liang et al., 2021] Jingyun Liang, Andreas Lugmayr, Kai Zhang, Martin Danelljan, Luc Van Gool, and Radu Tim- ofte. Hierarchical conditional flow: A unified framework for image super-resolution and image rescaling. In ICCV, pages 4076-4085, 2021.","title":"Hierarchical conditional flow: A unified framework for image super-resolution and image rescaling","authors":"Liang","year":"2021","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Lippe and Gavves, 2021] Phillip Lippe and Efstratios Gavves. Categorical normalizing flows via continuous transformations. In ICLR, 2021.","title":"Categorical normalizing flows via continuous transformations","authors":"Phillip Lippe, Efstratios Gavves","year":"2021","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Luo et al., 2021] Youzhi Luo, Keqiang Yan, and Shuiwang Ji. Graphdf: A discrete flow model for molecular graph generation. In ICML, 2021.","title":"Graphdf: A discrete flow model for molecular graph generation","authors":"Luo","year":"2021","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Madhawa et al., 2019] Kaushalya Madhawa, Katushiko Ishiguro, Kosuke Nakago, and Motoki Abe. Graphnvp: An invertible flow model for generating molecular graphs. arXiv preprint arXiv:1905.11600, 2019.","title":"Graphnvp: An invertible flow model for generating molecular graphs","authors":"Kaushalya Madhawa, Katushiko Ishiguro, Kosuke Nakago, Motoki Abe","year":"2019","venue":null,"doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Notin et al., 2021] Pascal Notin, José Miguel Hernández- Lobato, and Yarin Gal. Improving black-box optimization in vae latent space using decoder uncertainty. Advances in Neural Information Processing Systems, 34, 2021.","title":"Improving black-box optimization in vae latent space using decoder uncertainty","authors":"Pascal Notin, José Miguel Hernández-Lobato, Yarin Gal","year":"2021","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Papamakarios et al., 2017] George Papamakarios, Theo Pavlakou, and Iain Murray. Masked autoregressive flow for density estimation. Advances in neural information processing systems, 30, 2017.","title":"Masked autoregressive flow for density estimation","authors":"Papamakarios","year":"2017","venue":"Advances in neural information processing systems","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Papamakarios et al., 2021] George Papamakarios, Eric Nal- isnick, Danilo Jimenez Rezende, Shakir Mohamed, and Balaji Lakshminarayanan. Normalizing flows for proba- bilistic modeling and inference. JMLR, 2021.","title":"Normalizing flows for probabilistic modeling and inference","authors":"George Papamakarios, Eric Nalisnick, Danilo Jimenez Rezende, Shakir Mohamed, Balaji Lakshminarayanan","year":"2021","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Polishchuk et al., 2013] Pavel G Polishchuk, Timur I Madzhidov, and Alexandre Varnek. Estimation of the size of drug-like chemical space based on gdb-17 data. Jour- nal of computer-aided molecular design, 27(8):675-679, 2013.","title":"Estimation of the size of drug-like chemical space based on gdb-17 data","authors":"Polishchuk","year":"2013","venue":"Journal of computer-aided molecular design","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Ramachandran et al., 2017] Prajit Ramachandran, Barret Zoph, and Quoc V Le. Searching for activation functions. arXiv preprint arXiv:1710.05941, 2017.","title":"Searching for activation functions","authors":"Ramachandran","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Rezende and Mohamed, 2015] Danilo Rezende and Shakir Mohamed. Variational inference with normalizing flows. In ICML, 2015.","title":"Variational inference with normalizing flows","authors":"Danilo Rezende, Shakir Mohamed","year":"2015","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":30,"raw":"[Schlichtkrull et al., 2018] Michael Schlichtkrull, Thomas N Kipf, Peter Bloem, Rianne Van Den Berg, Ivan Titov, and Max Welling. Modeling relational data with graph convo- lutional networks. In ESWC, 2018.","title":"Modeling relational data with graph convolutional networks","authors":"Michael Schlichtkrull, Thomas N Kipf, Peter Bloem, Rianne Van Den, Ivan Berg, Max Titov, Welling","year":"2018","venue":"ESWC","doi":null,"url":null,"internal_url":null},{"position":31,"raw":"[Shi et al., 2020] Chence Shi, Minkai Xu, Zhaocheng Zhu, Weinan Zhang, Ming Zhang, and Jian Tang. Graphaf: a flow-based autoregressive model for molecular graph gen- eration. In ICLR, 2020.","title":"Graphaf: a flow-based autoregressive model for molecular graph generation","authors":"Shi","year":"2020","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":32,"raw":"[Simonovsky and Komodakis, 2018] Martin Simonovsky and Nikos Komodakis. Graphvae: Towards generation of small graphs using variational autoencoders. In ICANN. Springer, 2018.","title":"Graphvae: Towards generation of small graphs using variational autoencoders","authors":"Martin Simonovsky, Nikos Komodakis","year":"2018","venue":"ICANN","doi":null,"url":null,"internal_url":null},{"position":33,"raw":"[St. John et al., 2019] Peter C St. John, Caleb Phillips, et al. Message-passing neural networks for high-throughput polymer screening. The Journal of chemical physics, 150(23):234111, 2019.","title":"Message-passing neural networks for high-throughput polymer screening","authors":"C Peter, St, Caleb John, Phillips","year":"2019","venue":"The Journal of chemical physics","doi":null,"url":null,"internal_url":null},{"position":34,"raw":"[Tripp et al., 2020] Austin Tripp, Erik Daxberger, and José Miguel Hernández-Lobato. Sample-efficient opti- mization in the latent space of deep generative models via weighted retraining. Advances in Neural Information Pro- cessing Systems, 33:11259-11272, 2020.","title":"Sample-efficient optimization in the latent space of deep generative models via weighted retraining","authors":"Tripp","year":"2020","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":35,"raw":"[Xie et al., 2021] Yutong Xie, Chence Shi, Hao Zhou, Yuwei Yang, Weinan Zhang, Yong Yu, and Lei Li. Mars: Markov molecular sampling for multi-objective drug discovery. In ICLR, 2021.","title":"Mars: Markov molecular sampling for multi-objective drug discovery","authors":"Yutong Xie, Chence Shi, Hao Zhou, Yuwei Yang, Weinan Zhang, Yong Yu, Lei Li","year":"2021","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":36,"raw":"[Ying et al., 2018] Rex Ying, Jiaxuan You, Christopher Mor- ris, Xiang Ren, William L Hamilton, and Jure Leskovec. Hierarchical graph representation learning with differen- tiable pooling. In NeurIPS, 2018.","title":"Hierarchical graph representation learning with differentiable pooling","authors":"Rex Ying, Jiaxuan You, Christopher Morris, Xiang Ren, William L Hamilton, Jure Leskovec","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":37,"raw":"[You et al., 2018] Jiaxuan You, Bowen Liu, Rex Ying, Vi- jay Pande, and Jure Leskovec. Graph convolutional policy network for goal-directed molecular graph generation. In NeurIPS, 2018.","title":"Graph convolutional policy network for goal-directed molecular graph generation","authors":"Jiaxuan You, Bowen Liu, Rex Ying, Vijay Pande, Jure Leskovec","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":38,"raw":"[Yu et al., 2020] Jason J Yu, Konstantinos G Derpanis, and Marcus A Brubaker. Wavelet flow: Fast training of high resolution normalizing flows. NeurIPS, 33:6184-6196, 2020.","title":"Wavelet flow: Fast training of high resolution normalizing flows","authors":"Jason J Yu, Konstantinos G Derpanis, Marcus A Brubaker","year":"2020","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":39,"raw":"[Zang and Wang, 2020] Chengxi Zang and Fei Wang. Moflow: an invertible flow model for generating molecu- lar graphs. In SIGKDD, 2020.","title":"Moflow: an invertible flow model for generating molecular graphs","authors":"Wang ; ] Chengxi Zang, Fei Zang, Wang","year":"2020","venue":"SIGKDD","doi":null,"url":null,"internal_url":null},{"position":40,"raw":"[Zhu et al., 2023] Yiheng Zhu, Jialu Wu, Chaowen Hu, Ji- ahuan Yan, Chang-Yu Hsieh, Tingjun Hou, and Jian Wu. Sample-efficient multi-objective molecular optimization with gflownets. arXiv preprint arXiv:2302.04040, 2023.","title":"Sample-efficient multi-objective molecular optimization with gflownets","authors":"Yiheng Zhu, Jialu Wu, Chaowen Hu, Jiahuan Yan, Chang-Yu Hsieh, Tingjun Hou, Jian Wu","year":"2023","venue":null,"doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}