De novo generation of dual-target compounds using artificial intelligence

Kasumi Yasuda; Francois Berenger; Kazuma Amaike; Ayaka Ueda; Tomoya Nakagomi; Genki Hamasaki; Chen Li; Noriko Yuyama Otani; Kazuma Kaitoh; Koji Tsuda; Kenichiro Itami; Yoshihiro Yamanishi

doi:10.1016/j.isci.2024.111526

De novo generation of dual-target compounds using artificial intelligence

iScience. 2024 Dec 17;28(1):111526. doi: 10.1016/j.isci.2024.111526. eCollection 2025 Jan 17.

Authors

Kasumi Yasuda¹, Francois Berenger^{1

2}, Kazuma Amaike³, Ayaka Ueda³, Tomoya Nakagomi³, Genki Hamasaki¹, Chen Li^{1

4}, Noriko Yuyama Otani^{1

4}, Kazuma Kaitoh^{1

4}, Koji Tsuda², Kenichiro Itami^{3

5}, Yoshihiro Yamanishi^{1

4}

Affiliations

¹ Department of Bioscience and Bioinformatics, Kyushu Institute of Technology, 680-4 Kawazu, Iizuka, Fukuoka 820-8502, Japan.
² Graduate School of Frontier Sciences, The University of Tokyo, 5-1-5 Kashiwa-no-ha, Kashiwa, Chiba 277-8561, Japan.
³ Graduate School of Science, Nagoya University, Chikusa, Nagoya, Aichi 464-8602, Japan.
⁴ Graduate School of Informatics, Nagoya University, Chikusa, Nagoya, Aichi 464-8601, Japan.
⁵ Institute of Transformative Bio-Molecules (WPI-ITbM), Nagoya University, Chikusa, Nagoya, Aichi 464-8602, Japan.

Abstract

Drugs that interact with multiple therapeutic targets are potential high-value products in polypharmacology-based drug discovery, but the rational design remains a formidable challenge. Here, we present artificial intelligence (AI)-based methods to design the chemical structures of compounds that interact with multiple therapeutic target proteins. The molecular structure generation is performed by a fragment-based approach using a genetic algorithm with chemical substructures and a deep learning approach using reinforcement learning with stochastic policy gradients in the framework of generative adversarial networks. Using the proposed methods, we designed the chemical structures of compounds that would interact with two therapeutic targets of bronchial asthma, i.e., adenosine A2a receptor (ADORA2A) and phosphodiesterase 4D (PDE4D). We then synthesized 10 compounds and evaluated their bioactivities via the binding assays of 39 target human proteins, including ADORA2A and PDE4D. Three of the 10 synthesized compounds successfully interacted with ADORA2A and PDE4D with high specificity.

Keywords: Bioinformatics; Biological sciences; Natural sciences; Pharmacoinformatics.