{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deep-reinforcement-learning-for-de-novo-drug","title":"Deep Reinforcement Learning for De-Novo Drug Design","arxiv_id":"1711.10907","date":"2017-11-29","proceeding":null,"authors":["Mariya Popova","Olexandr Isayev","Alexander Tropsha"],"abstract":"We propose a novel computational strategy for de novo design of molecules\nwith desired properties termed ReLeaSE (Reinforcement Learning for Structural\nEvolution). Based on deep and reinforcement learning approaches, ReLeaSE\nintegrates two deep neural networks - generative and predictive - that are\ntrained separately but employed jointly to generate novel targeted chemical\nlibraries. ReLeaSE employs simple representation of molecules by their SMILES\nstrings only. Generative models are trained with stack-augmented memory network\nto produce chemically feasible SMILES strings, and predictive models are\nderived to forecast the desired properties of the de novo generated compounds.\nIn the first phase of the method, generative and predictive models are trained\nseparately with a supervised learning algorithm. In the second phase, both\nmodels are trained jointly with the reinforcement learning approach to bias the\ngeneration of new chemical structures towards those with the desired physical\nand/or biological properties. In the proof-of-concept study, we have employed\nthe ReLeaSE method to design chemical libraries with a bias toward structural\ncomplexity or biased toward compounds with either maximal, minimal, or specific\nrange of physical properties such as melting point or hydrophobicity, as well\nas to develop novel putative inhibitors of JAK2. The approach proposed herein\ncan find a general use for generating targeted chemical libraries of novel\ncompounds optimized for either a single desired property or multiple\nproperties.","url_abs":"http://arxiv.org/abs/1711.10907v2","url_pdf":"http://arxiv.org/pdf/1711.10907v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deep-reinforcement-learning-for-de-novo-drug","repo_url":"https://github.com/isayev/ReLeaSE","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"deep-reinforcement-learning","task_name":"Deep Reinforcement Learning"},{"task_slug":"drug-design","task_name":"Drug Design"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1711.10907","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}