AB-Gen:利用生成式预训练变换器和深度强化学习设计抗体库。

IF 11.5 2区 生物学 Q1 GENETICS & HEREDITY Genomics, Proteomics & Bioinformatics Pub Date : 2023-10-01 Epub Date: 2023-06-24 DOI:10.1016/j.gpb.2023.03.004
Xiaopeng Xu, Tiantian Xu, Juexiao Zhou, Xingyu Liao, Ruochi Zhang, Yu Wang, Lu Zhang, Xin Gao
{"title":"AB-Gen:利用生成式预训练变换器和深度强化学习设计抗体库。","authors":"Xiaopeng Xu, Tiantian Xu, Juexiao Zhou, Xingyu Liao, Ruochi Zhang, Yu Wang, Lu Zhang, Xin Gao","doi":"10.1016/j.gpb.2023.03.004","DOIUrl":null,"url":null,"abstract":"<p><p>Antibody leads must fulfill multiple desirable properties to be clinical candidates. Primarily due to the low throughput in the experimental procedure, the need for such multi-property optimization causes the bottleneck in preclinical antibody discovery and development, because addressing one issue usually causes another. We developed a reinforcement learning (RL) method, named AB-Gen, for antibody library design using a generative pre-trained transformer (GPT) as the policy network of the RL agent. We showed that this model can learn the antibody space of heavy chain complementarity determining region 3 (CDRH3) and generate sequences with similar property distributions. Besides, when using human epidermal growth factor receptor-2 (HER2) as the target, the agent model of AB-Gen was able to generate novel CDRH3 sequences that fulfill multi-property constraints. Totally, 509 generated sequences were able to pass all property filters, and three highly conserved residues were identified. The importance of these residues was further demonstrated by molecular dynamics simulations, consolidating that the agent model was capable of grasping important information in this complex optimization task. Overall, the AB-Gen method is able to design novel antibody sequences with an improved success rate than the traditional propose-then-filter approach. It has the potential to be used in practical antibody design, thus empowering the antibody discovery and development process. The source code of AB-Gen is freely available at Zenodo (https://doi.org/10.5281/zenodo.7657016) and BioCode (https://ngdc.cncb.ac.cn/biocode/tools/BT007341).</p>","PeriodicalId":12528,"journal":{"name":"Genomics, Proteomics & Bioinformatics","volume":null,"pages":null},"PeriodicalIF":11.5000,"publicationDate":"2023-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10928431/pdf/","citationCount":"0","resultStr":"{\"title\":\"AB-Gen: Antibody Library Design with Generative Pre-trained Transformer and Deep Reinforcement Learning.\",\"authors\":\"Xiaopeng Xu, Tiantian Xu, Juexiao Zhou, Xingyu Liao, Ruochi Zhang, Yu Wang, Lu Zhang, Xin Gao\",\"doi\":\"10.1016/j.gpb.2023.03.004\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>Antibody leads must fulfill multiple desirable properties to be clinical candidates. Primarily due to the low throughput in the experimental procedure, the need for such multi-property optimization causes the bottleneck in preclinical antibody discovery and development, because addressing one issue usually causes another. We developed a reinforcement learning (RL) method, named AB-Gen, for antibody library design using a generative pre-trained transformer (GPT) as the policy network of the RL agent. We showed that this model can learn the antibody space of heavy chain complementarity determining region 3 (CDRH3) and generate sequences with similar property distributions. Besides, when using human epidermal growth factor receptor-2 (HER2) as the target, the agent model of AB-Gen was able to generate novel CDRH3 sequences that fulfill multi-property constraints. Totally, 509 generated sequences were able to pass all property filters, and three highly conserved residues were identified. The importance of these residues was further demonstrated by molecular dynamics simulations, consolidating that the agent model was capable of grasping important information in this complex optimization task. Overall, the AB-Gen method is able to design novel antibody sequences with an improved success rate than the traditional propose-then-filter approach. It has the potential to be used in practical antibody design, thus empowering the antibody discovery and development process. The source code of AB-Gen is freely available at Zenodo (https://doi.org/10.5281/zenodo.7657016) and BioCode (https://ngdc.cncb.ac.cn/biocode/tools/BT007341).</p>\",\"PeriodicalId\":12528,\"journal\":{\"name\":\"Genomics, Proteomics & Bioinformatics\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":11.5000,\"publicationDate\":\"2023-10-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10928431/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Genomics, Proteomics & Bioinformatics\",\"FirstCategoryId\":\"99\",\"ListUrlMain\":\"https://doi.org/10.1016/j.gpb.2023.03.004\",\"RegionNum\":2,\"RegionCategory\":\"生物学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2023/6/24 0:00:00\",\"PubModel\":\"Epub\",\"JCR\":\"Q1\",\"JCRName\":\"GENETICS & HEREDITY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Genomics, Proteomics & Bioinformatics","FirstCategoryId":"99","ListUrlMain":"https://doi.org/10.1016/j.gpb.2023.03.004","RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2023/6/24 0:00:00","PubModel":"Epub","JCR":"Q1","JCRName":"GENETICS & HEREDITY","Score":null,"Total":0}
引用次数: 0

摘要

抗体先导物必须满足多种理想特性才能成为临床候选物。主要由于实验过程的吞吐量较低,这种多属性优化的需求造成了临床前抗体发现和开发的瓶颈,因为解决一个问题通常会引发另一个问题。我们开发了一种用于抗体库设计的强化学习(RL)方法,命名为 AB-Gen,使用生成式预训练变换器(GPT)作为 RL 代理的策略网络。我们的研究表明,该模型可以学习重链互补决定区 3(CDRH3)的抗体空间,并生成具有相似性质分布的序列。此外,当使用人表皮生长因子受体-2(HER2)作为靶点时,AB-Gen 的代理模型能够生成满足多属性约束的新型 CDRH3 序列。总共有 509 个生成的序列能够通过所有属性筛选,并确定了三个高度保守的残基。分子动力学模拟进一步证明了这些残基的重要性,从而巩固了代理模型能够在这项复杂的优化任务中掌握重要信息。总之,与传统的 "提出-然后过滤 "方法相比,AB-Gen 方法能够提高设计新型抗体序列的成功率。它有望用于实际的抗体设计,从而促进抗体的发现和开发过程。AB-Gen 的源代码可在 Zenodo (https://doi.org/10.5281/zenodo.7657016) 和 BioCode (https://ngdc.cncb.ac.cn/biocode/tools/BT007341) 免费获取。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
AB-Gen: Antibody Library Design with Generative Pre-trained Transformer and Deep Reinforcement Learning.

Antibody leads must fulfill multiple desirable properties to be clinical candidates. Primarily due to the low throughput in the experimental procedure, the need for such multi-property optimization causes the bottleneck in preclinical antibody discovery and development, because addressing one issue usually causes another. We developed a reinforcement learning (RL) method, named AB-Gen, for antibody library design using a generative pre-trained transformer (GPT) as the policy network of the RL agent. We showed that this model can learn the antibody space of heavy chain complementarity determining region 3 (CDRH3) and generate sequences with similar property distributions. Besides, when using human epidermal growth factor receptor-2 (HER2) as the target, the agent model of AB-Gen was able to generate novel CDRH3 sequences that fulfill multi-property constraints. Totally, 509 generated sequences were able to pass all property filters, and three highly conserved residues were identified. The importance of these residues was further demonstrated by molecular dynamics simulations, consolidating that the agent model was capable of grasping important information in this complex optimization task. Overall, the AB-Gen method is able to design novel antibody sequences with an improved success rate than the traditional propose-then-filter approach. It has the potential to be used in practical antibody design, thus empowering the antibody discovery and development process. The source code of AB-Gen is freely available at Zenodo (https://doi.org/10.5281/zenodo.7657016) and BioCode (https://ngdc.cncb.ac.cn/biocode/tools/BT007341).

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Genomics, Proteomics & Bioinformatics
Genomics, Proteomics & Bioinformatics Biochemistry, Genetics and Molecular Biology-Biochemistry
CiteScore
14.30
自引率
4.20%
发文量
844
审稿时长
61 days
期刊介绍: Genomics, Proteomics and Bioinformatics (GPB) is the official journal of the Beijing Institute of Genomics, Chinese Academy of Sciences / China National Center for Bioinformation and Genetics Society of China. It aims to disseminate new developments in the field of omics and bioinformatics, publish high-quality discoveries quickly, and promote open access and online publication. GPB welcomes submissions in all areas of life science, biology, and biomedicine, with a focus on large data acquisition, analysis, and curation. Manuscripts covering omics and related bioinformatics topics are particularly encouraged. GPB is indexed/abstracted by PubMed/MEDLINE, PubMed Central, Scopus, BIOSIS Previews, Chemical Abstracts, CSCD, among others.
期刊最新文献
Review and Evaluate the Bioinformatics Analysis Strategies of ATAC-seq and CUT&Tag Data. Identification of highly repetitive barley enhancers with long-range regulation potential via STARR-seq CpG island definition and methylation mapping of the T2T-YAO genome Pindel-TD: a tandem duplication detector based on a pattern growth approach SMARTdb: An Integrated Database for Exploring Single-cell Multi-omics Data of Reproductive Medicine
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1