PCW-A1001, AI-assisted de novo design approach to design a selective inhibitor for FLT-3(D835Y) in acute myeloid leukemia

Jang, Seong Hun; Sivakumar, Dakshinamurthy; Mudedla, Sathish Kumar; Choi, Jaehan; Lee, Sung-Min; Jeon, Minjun; Bvs, Suneel Kumar; Hwang, Jinha; Kang, Minsung; Shin, Eun Gyeong; Lee, Kyu Myung; Jung, Kwan‐Young; Kim, Jae-Sung; Wu, Sangwook

doi:10.3389/fmolb.2022.1072028

Cited by 6 publications

(5 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[Public] February 9, 2024 generation, music generation, art generation, and also for drug discovery. [30][31][32][33] A filtered ChEMBL30 dataset, a repository rich in chemical diversity, was used to train the RNN-LSTM model initially. We trained our RNN-LSTM model on SMILES (Simplified Molecular Input Line Entry System), a textual representation of molecular structures, to accurately capture the chemical data.…”

Section: Ai-guided Fragment Expansionmentioning

confidence: 99%

The Third CACHE Challenge – Finding Ligands Targeting the Macrodomain of SARS-CoV-2 NSP3 Using AI-inspired and Knowledge-Based Approaches.

Kumar B.V.S,

Rostaing,

Burai-Patrascu

et al. 2024

Preprint

View full text Add to dashboard Cite

The SARS-CoV-2 virus contains a host of nonstructural proteins (NSPs) that contribute to its structure and viral function. Among them is the nonstructural protein 3 (NSP3), which contains a macrodomain (Mac1) that interferes with antiviral adenosine diphosphate (ADP)-ribosylation signaling. Catalytic mutations in Mac1 render viruses nonpathogenic, making this enzyme a promising target for antiviral development. For this reason, the third CACHE challenge focused on identifying binders of the Mac1 domain of NSP3 for the development of novel antivirals against SARS-CoV-2. To this end, we used available structural data of the NSP3 Mac1 domain in complex with known fragment binders as starting points for ligand discovery; our efforts were primarily focused on sub-sites of the ADP binding site in the NSP3 macrodomain. Then, using Artificial intelligence (AI)-guided and knowledge-based fragment merging and expansion approaches, we generated novel molecules that would serve as templates to identify highly similar compounds in the Enamine REAL database that would be commercially available. Our design yielded a library of 12,800 molecules, which was docked with our program FITTED to a representative crystal structure of NSP3. We ranked the predicted binding poses based on docking score, followed by visual pose analysis of the best 200 compounds. We finally selected and proposed 150 compounds for testing, followed by further shortlisting to yield a final list of 107 molecules. 91 compounds were purchased from Enamine and are being tested at the Structural Genomics Consortium (SGC). Our approach and findings will further contribute to our open science efforts, and we aim to continue to engage the scientific community.

show abstract

Section: Ai-guided Fragment Expansionmentioning

confidence: 99%

The Third CACHE Challenge – Finding Ligands Targeting the Macrodomain of SARS-CoV-2 NSP3 Using AI-inspired and Knowledge-Based Approaches.

Kumar B.V.S,

Rostaing,

Burai-Patrascu

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…AI has already produced successful drug development cases in various therapeutic areas, such as the A2A receptor antagonist EXS-21546 created in just eight months by Evotec and Excientia. Similarly, Insilico Medicine's AI program produced a new inhibitor (ISM001-055) for antifibrotic targets in nine months [12]. Despite these remarkable achievements, AI has yet to be widely applied to TCM, an ancient medical practice used for thousands of years and holds a wealth of knowledge and experience [13,14].…”

Section: Introductionmentioning

confidence: 99%

The Application of Artificial Intelligence in the Research and Development of Traditional Chinese Medicine

Ke,

Liu,

Liu

et al. 2024

IJDDP

View full text Add to dashboard Cite

Review The Application of Artificial Intelligence in the Research and Development of Traditional Chinese Medicine Zhipeng Ke 1,2, Minxuan Liu 1,2,3, Jing Liu 1,2, Zhenzhen Su 1,2, Lu Li 1,2, Mengyu Qian 1,2, Xinzhuang Zhang 1,2, Tuanjie Wang 1,2, Liang Cao 1,2, Zhenzhong Wang 1,2, and Wei Xiao 1,2, * 1 National Key Laboratory on Technologies for Chinese Medicine Pharmaceutical Process Control and Intelligent Manufacture, Lianyungang 222106, China 2 Jiangsu Kanion Pharmaceutical Co., Ltd, Lianyungang 222104, China 3 ‍School of Pharmacy, Nanjing University of Chinese Medicine, Nanjing 210009, China * Correspondence: xw_kanion@163.com Received: 4 September 2023 Accepted: 4 November 2023 Published: 6 March 2024 Abstract: With the accumulation of data in the pharmaceutical industry and the development of artificial intelligence technology, various artificial intelligence methods have been successfully employed in the drug discovery process. The integration of artificial intelligence in Traditional Chinese medicine has also gained momentum, encompassing quality control of Chinese patent medicines, prescriptions optimization, discovery of effective substances, and prediction of side effects. However, artificial intelligence also faces challenges and limitations in Traditional Chinese medicine development, such as data scarcity and complexity, lack of interdisciplinary professionals, black-box models, etc. Therefore, more research and collaboration are needed to address these issues and explore the best ways to integrate artificial intelligence and Traditional Chinese medicine to improve human health.

show abstract

“…Recent advancements in de novo molecular design have positioned generative methods as a complementary approach to traditional virtual screening. ,− Core advantages of these models include the ability to sample chemical space outside the training data and by coupling an optimization algorithm, goal-directed learning can be achieved . Although the field is relatively nascent, molecular generative models have identified experimentally validated therapeutic molecules ,,− and organocatalysts . An important shared commonality between these success stories is the inclusion of relatively computationally expensive oracles.…”

Section: Introductionmentioning

confidence: 99%

Augmented Memory: Sample-Efficient Generative Molecular Design with Reinforcement Learning

Guo,

Schwaller

2024

JACS Au

View full text Add to dashboard Cite

Sample efficiency is a fundamental challenge in de novo molecular design. Ideally, molecular generative models should learn to satisfy a desired objective under minimal calls to oracles (computational property predictors). This problem becomes more apparent when using oracles that can provide increased predictive accuracy but impose significant computational cost. Consequently, designing molecules that are optimized for such oracles cannot be achieved under a practical computational budget. Molecular generative models based on simplified molecular-input line-entry system (SMILES) have shown remarkable sample efficiency when coupled with reinforcement learning, as demonstrated in the practical molecular optimization (PMO) benchmark. Here, we first show that experience replay drastically improves the performance of multiple previously proposed algorithms. Next, we propose a novel algorithm called Augmented Memory that combines data augmentation with experience replay. We show that scores obtained from oracle calls can be reused to update the model multiple times. We compare Augmented Memory to previously proposed algorithms and show significantly enhanced sample efficiency in an exploitation task, a drug discovery case study requiring both exploration and exploitation, and a materials design case study optimizing explicitly for quantummechanical properties. Our method achieves a new state-of-the-art in sample-efficient de novo molecular design, outperforming all of the previously reported methods. The code is available at https://github.com/schwallergroup/augmented_memory.

show abstract

PCW-A1001, AI-assisted de novo design approach to design a selective inhibitor for FLT-3(D835Y) in acute myeloid leukemia

Cited by 6 publications

References 69 publications

The Third CACHE Challenge – Finding Ligands Targeting the Macrodomain of SARS-CoV-2 NSP3 Using AI-inspired and Knowledge-Based Approaches.

The Third CACHE Challenge – Finding Ligands Targeting the Macrodomain of SARS-CoV-2 NSP3 Using AI-inspired and Knowledge-Based Approaches.

The Application of Artificial Intelligence in the Research and Development of Traditional Chinese Medicine

Augmented Memory: Sample-Efficient Generative Molecular Design with Reinforcement Learning

Contact Info

Product

Resources

About