The objective of drug discovery is to identify chemical compounds that possess specific pharmaceutical properties toward a binding target. Existing large language models (LLMS) can achieve high token matching scores in terms of likelihood for molecule generation. However, relying solely on LLM decoding often results in the generation of molecules that are either invalid due to a single misused token, or suboptimal due to unbalanced exploration and exploitation as a consequence of the LLMs prior experience. Here we propose ERP, Entropy-Reinforced Planning for Transformer Decoding, which employs an entropy-reinforced planning algorithm to enhance the Transformer decoding process and strike a balance between exploitation and exploration. ERP aims to achieve improvements in multiple properties compared to direct sampling from the Transformer. We evaluated ERP on the SARS-CoV-2 virus (3CLPro) and human cancer cell target protein (RTCB) benchmarks and demonstrated that, in both benchmarks, ERP consistently outperforms the current state-of-the-art algorithm by 1-5 percent, and baselines by 5-10 percent, respectively. Moreover, such improvement is robust across Transformer models trained with different objectives. Finally, to further illustrate the capabilities of ERP, we tested our algorithm on three code generation benchmarks and outperformed the current state-of-the-art approach as well. Our code is publicly available at: https://github.com/xuefeng-cs/ERP.
Fast and scalable retrosynthetic planning with a transformer neural network and speculative beam search
SUPSI, IDSIA(SUPSI、IDSIA);Pfizer Research and Development(辉瑞研发);Institute for Digital Technologies for Personalized Healthcare(个性化医疗数字技术研究所);KAUST(王国立科技大学);Center for Generative AI(生成人工智能中心)
URSA: Chemistry-Aware Benchmark for Utilitarian Retrosynthesis Assessment
Insilico Medicine AI Limited(Insilico Medicine AI有限公司);Independent researcher(独立研究者);Insilico Medicine Canada Inc.(Insilico Medicine加拿大公司);Insilico Medicine Hong Kong Ltd.(Insilico Medicine香港有限公司)
AOT*: Efficient Synthesis Planning via LLM-Empowered AND-OR Tree Search
School of Data Science, CUHK-Shenzhen(CUHK-Shenzhen数据科学学院);Shanghai AI Laboratory(上海人工智能实验室);Jilin University(吉林大学);Mila - Quebec AI Institute(魁北克AI研究所);HEC Montréal(蒙特利尔HEC)
LLM-Augmented Chemical Synthesis and Design Decision Programs
École Polytechnique Fédérale de Lausanne (EPFL)(瑞士联邦理工学院洛桑校区);National Centre of Competence in Research (NCCR) Catalysis(催化领域竞争力研究国立中心);Harvard University(哈佛大学);Cornell University(康奈尔大学)
RetroDFM-R: Reasoning-Driven Retrosynthesis Prediction with Large Language Models via Reinforcement Learning
X-LANCE Lab, School of Computer Science MoE Key Lab of Artificial Intelligence SJTU AI Institute(X-LANCE实验室,计算机科学学院,人工智能关键实验室,SJTU人工智能研究所);Suzhou Laboratory(苏州实验室);Jiangsu Key Lab of Language Computing(江苏语言计算重点实验室);School of Chemistry and Chemical Engineering(化学与化工学院)
CombiMOTS: Combinatorial Multi-Objective Tree Search for Dual-Target Molecule Generation
Department of Computer Science and Engineering, Seoul National University, Seoul, Republic of Korea(首尔国立大学计算机科学与工程系);Interdisciplinary Program in Bioinformatics, Seoul National University, Seoul, Republic of Korea(首尔国立大学生物信息学跨学科项目);AIGENDRUG Co., Ltd., Seoul, Republic of Korea(AIGENDRUG公司);Interdisciplinary Program in Artificial Intelligence, Seoul National University, Seoul, Republic of Korea(首尔国立大学人工智能跨学科项目)
Language Interaction Network for Clinical Trial Approval Estimation
University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);Rensselaer Polytechnic Institute(伦斯勒理工学院);Carle Illinois College of Medicine, University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校卡尔伊利诺伊医学院)