Yao Pei, Marcus Ho-Hin Shum, Yunshi Liao, Vivian W Leung, Yu-Nong Gong, David K Smith, Xiaole Yin, Yi Guan, Ruibang Luo, Tong Zhang, Tommy Tsan-Yuk Lam
{"title":"ARGNet:使用深度神经网络从序列中对抗生素耐药基因进行稳健识别和分类。","authors":"Yao Pei, Marcus Ho-Hin Shum, Yunshi Liao, Vivian W Leung, Yu-Nong Gong, David K Smith, Xiaole Yin, Yi Guan, Ruibang Luo, Tong Zhang, Tommy Tsan-Yuk Lam","doi":"10.1186/s40168-024-01805-0","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Emergence of antibiotic resistance in bacteria is an important threat to global health. Antibiotic resistance genes (ARGs) are some of the key components to define bacterial resistance and their spread in different environments. Identification of ARGs, particularly from high-throughput sequencing data of the specimens, is the state-of-the-art method for comprehensively monitoring their spread and evolution. Current computational methods to identify ARGs mainly rely on alignment-based sequence similarities with known ARGs. Such approaches are limited by choice of reference databases and may potentially miss novel ARGs. The similarity thresholds are usually simple and could not accommodate variations across different gene families and regions. It is also difficult to scale up when sequence data are increasing.</p><p><strong>Results: </strong>In this study, we developed ARGNet, a deep neural network that incorporates an unsupervised learning autoencoder model to identify ARGs and a multiclass classification convolutional neural network to classify ARGs that do not depend on sequence alignment. This approach enables a more efficient discovery of both known and novel ARGs. ARGNet accepts both amino acid and nucleotide sequences of variable lengths, from partial (30-50 aa; 100-150 nt) sequences to full-length protein or genes, allowing its application in both target sequencing and metagenomic sequencing. Our performance evaluation showed that ARGNet outperformed other deep learning models including DeepARG and HMD-ARG in most of the application scenarios especially quasi-negative test and the analysis of prediction consistency with phylogenetic tree. ARGNet has a reduced inference runtime by up to 57% relative to DeepARG.</p><p><strong>Conclusions: </strong>ARGNet is flexible, efficient, and accurate at predicting a broad range of ARGs from the sequencing data. ARGNet is freely available at https://github.com/id-bioinfo/ARGNet , with an online service provided at https://ARGNet.hku.hk . Video Abstract.</p>","PeriodicalId":18447,"journal":{"name":"Microbiome","volume":null,"pages":null},"PeriodicalIF":13.8000,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11080312/pdf/","citationCount":"0","resultStr":"{\"title\":\"ARGNet: using deep neural networks for robust identification and classification of antibiotic resistance genes from sequences.\",\"authors\":\"Yao Pei, Marcus Ho-Hin Shum, Yunshi Liao, Vivian W Leung, Yu-Nong Gong, David K Smith, Xiaole Yin, Yi Guan, Ruibang Luo, Tong Zhang, Tommy Tsan-Yuk Lam\",\"doi\":\"10.1186/s40168-024-01805-0\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Background: </strong>Emergence of antibiotic resistance in bacteria is an important threat to global health. Antibiotic resistance genes (ARGs) are some of the key components to define bacterial resistance and their spread in different environments. Identification of ARGs, particularly from high-throughput sequencing data of the specimens, is the state-of-the-art method for comprehensively monitoring their spread and evolution. Current computational methods to identify ARGs mainly rely on alignment-based sequence similarities with known ARGs. Such approaches are limited by choice of reference databases and may potentially miss novel ARGs. The similarity thresholds are usually simple and could not accommodate variations across different gene families and regions. It is also difficult to scale up when sequence data are increasing.</p><p><strong>Results: </strong>In this study, we developed ARGNet, a deep neural network that incorporates an unsupervised learning autoencoder model to identify ARGs and a multiclass classification convolutional neural network to classify ARGs that do not depend on sequence alignment. This approach enables a more efficient discovery of both known and novel ARGs. ARGNet accepts both amino acid and nucleotide sequences of variable lengths, from partial (30-50 aa; 100-150 nt) sequences to full-length protein or genes, allowing its application in both target sequencing and metagenomic sequencing. Our performance evaluation showed that ARGNet outperformed other deep learning models including DeepARG and HMD-ARG in most of the application scenarios especially quasi-negative test and the analysis of prediction consistency with phylogenetic tree. ARGNet has a reduced inference runtime by up to 57% relative to DeepARG.</p><p><strong>Conclusions: </strong>ARGNet is flexible, efficient, and accurate at predicting a broad range of ARGs from the sequencing data. ARGNet is freely available at https://github.com/id-bioinfo/ARGNet , with an online service provided at https://ARGNet.hku.hk . Video Abstract.</p>\",\"PeriodicalId\":18447,\"journal\":{\"name\":\"Microbiome\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":13.8000,\"publicationDate\":\"2024-05-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11080312/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Microbiome\",\"FirstCategoryId\":\"99\",\"ListUrlMain\":\"https://doi.org/10.1186/s40168-024-01805-0\",\"RegionNum\":1,\"RegionCategory\":\"生物学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"MICROBIOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Microbiome","FirstCategoryId":"99","ListUrlMain":"https://doi.org/10.1186/s40168-024-01805-0","RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MICROBIOLOGY","Score":null,"Total":0}
ARGNet: using deep neural networks for robust identification and classification of antibiotic resistance genes from sequences.
Background: Emergence of antibiotic resistance in bacteria is an important threat to global health. Antibiotic resistance genes (ARGs) are some of the key components to define bacterial resistance and their spread in different environments. Identification of ARGs, particularly from high-throughput sequencing data of the specimens, is the state-of-the-art method for comprehensively monitoring their spread and evolution. Current computational methods to identify ARGs mainly rely on alignment-based sequence similarities with known ARGs. Such approaches are limited by choice of reference databases and may potentially miss novel ARGs. The similarity thresholds are usually simple and could not accommodate variations across different gene families and regions. It is also difficult to scale up when sequence data are increasing.
Results: In this study, we developed ARGNet, a deep neural network that incorporates an unsupervised learning autoencoder model to identify ARGs and a multiclass classification convolutional neural network to classify ARGs that do not depend on sequence alignment. This approach enables a more efficient discovery of both known and novel ARGs. ARGNet accepts both amino acid and nucleotide sequences of variable lengths, from partial (30-50 aa; 100-150 nt) sequences to full-length protein or genes, allowing its application in both target sequencing and metagenomic sequencing. Our performance evaluation showed that ARGNet outperformed other deep learning models including DeepARG and HMD-ARG in most of the application scenarios especially quasi-negative test and the analysis of prediction consistency with phylogenetic tree. ARGNet has a reduced inference runtime by up to 57% relative to DeepARG.
Conclusions: ARGNet is flexible, efficient, and accurate at predicting a broad range of ARGs from the sequencing data. ARGNet is freely available at https://github.com/id-bioinfo/ARGNet , with an online service provided at https://ARGNet.hku.hk . Video Abstract.
期刊介绍:
Microbiome is a journal that focuses on studies of microbiomes in humans, animals, plants, and the environment. It covers both natural and manipulated microbiomes, such as those in agriculture. The journal is interested in research that uses meta-omics approaches or novel bioinformatics tools and emphasizes the community/host interaction and structure-function relationship within the microbiome. Studies that go beyond descriptive omics surveys and include experimental or theoretical approaches will be considered for publication. The journal also encourages research that establishes cause and effect relationships and supports proposed microbiome functions. However, studies of individual microbial isolates/species without exploring their impact on the host or the complex microbiome structures and functions will not be considered for publication. Microbiome is indexed in BIOSIS, Current Contents, DOAJ, Embase, MEDLINE, PubMed, PubMed Central, and Science Citations Index Expanded.