63
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      DeepARG: a deep learning approach for predicting antibiotic resistance genes from metagenomic data

      research-article

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Background

          Growing concerns about increasing rates of antibiotic resistance call for expanded and comprehensive global monitoring. Advancing methods for monitoring of environmental media (e.g., wastewater, agricultural waste, food, and water) is especially needed for identifying potential resources of novel antibiotic resistance genes (ARGs), hot spots for gene exchange, and as pathways for the spread of ARGs and human exposure. Next-generation sequencing now enables direct access and profiling of the total metagenomic DNA pool, where ARGs are typically identified or predicted based on the “best hits” of sequence searches against existing databases. Unfortunately, this approach produces a high rate of false negatives. To address such limitations, we propose here a deep learning approach, taking into account a dissimilarity matrix created using all known categories of ARGs. Two deep learning models, DeepARG-SS and DeepARG-LS, were constructed for short read sequences and full gene length sequences, respectively.

          Results

          Evaluation of the deep learning models over 30 antibiotic resistance categories demonstrates that the DeepARG models can predict ARGs with both high precision (> 0.97) and recall (> 0.90). The models displayed an advantage over the typical best hit approach, yielding consistently lower false negative rates and thus higher overall recall (> 0.9). As more data become available for under-represented ARG categories, the DeepARG models’ performance can be expected to be further enhanced due to the nature of the underlying neural networks. Our newly developed ARG database, DeepARG-DB, encompasses ARGs predicted with a high degree of confidence and extensive manual inspection, greatly expanding current ARG repositories.

          Conclusions

          The deep learning models developed here offer more accurate antimicrobial resistance annotation relative to current bioinformatics practice. DeepARG does not require strict cutoffs, which enables identification of a much broader diversity of ARGs. The DeepARG models and database are available as a command line version and as a Web service at http://bench.cs.vt.edu/deeparg.

          Electronic supplementary material

          The online version of this article (10.1186/s40168-018-0401-z) contains supplementary material, which is available to authorized users.

          Related collections

          Most cited references44

          • Record: found
          • Abstract: found
          • Article: not found

          Tackling antibiotic resistance: the environmental framework.

          Antibiotic resistance is a threat to human and animal health worldwide, and key measures are required to reduce the risks posed by antibiotic resistance genes that occur in the environment. These measures include the identification of critical points of control, the development of reliable surveillance and risk assessment procedures, and the implementation of technological solutions that can prevent environmental contamination with antibiotic resistant bacteria and genes. In this Opinion article, we discuss the main knowledge gaps, the future research needs and the policy and management options that should be prioritized to tackle antibiotic resistance in the environment.
            Bookmark
            • Record: found
            • Abstract: found
            • Article: found
            Is Open Access

            ARDB—Antibiotic Resistance Genes Database

            The treatment of infections is increasingly compromised by the ability of bacteria to develop resistance to antibiotics through mutations or through the acquisition of resistance genes. Antibiotic resistance genes also have the potential to be used for bio-terror purposes through genetically modified organisms. In order to facilitate the identification and characterization of these genes, we have created a manually curated database—the Antibiotic Resistance Genes Database (ARDB)—unifying most of the publicly available information on antibiotic resistance. Each gene and resistance type is annotated with rich information, including resistance profile, mechanism of action, ontology, COG and CDD annotations, as well as external links to sequence and protein databases. Our database also supports sequence similarity searches and implements an initial version of a tool for characterizing common mutations that confer antibiotic resistance. The information we provide can be used as compendium of antibiotic resistance factors as well as to identify the resistance genes of newly sequenced genes, genomes, or metagenomes. Currently, ARDB contains resistance information for 13 293 genes, 377 types, 257 antibiotics, 632 genomes, 933 species and 124 genera. ARDB is available at http://ardb.cbcb.umd.edu/.
              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Bacterial phylogeny structures soil resistomes across habitats

              Summary Ancient and diverse antibiotic resistance genes (ARGs) have previously been identified from soil 1–3 , including genes identical to those in human pathogens 4 . Despite the apparent overlap between soil and clinical resistomes 4–6 , factors influencing ARG composition in soil and their movement between genomes and habitats remain largely unknown 3 . General metagenome functions often correlate with the underlying structure of bacterial communities 7–12 . However, ARGs are hypothesized to be highly mobile 4,5,13 , prompting speculation that resistomes may not correlate with phylogenetic signatures or ecological divisions 13,14 . To investigate these relationships, we performed functional metagenomic selections for resistance to 18 antibiotics from 18 agricultural and grassland soils. The 2895 ARGs we discovered were predominantly novel, and represent all major resistance mechanisms 15 . We demonstrate that distinct soil types harbor distinct resistomes, and that nitrogen fertilizer amendments strongly influenced soil ARG content. Resistome composition also correlated with microbial phylogenetic and taxonomic structure, both across and within soil types. Consistent with this strong correlation, mobility elements syntenic with ARGs were rare in soil compared to sequenced pathogens, suggesting that ARGs in the soil may not transfer between bacteria as readily as is observed in the clinic. Together, our results indicate that bacterial community composition is the primary determinant of soil ARG content, challenging previous hypotheses that horizontal gene transfer effectively decouples resistomes from phylogeny 13,14 .
                Bookmark

                Author and article information

                Contributors
                gustavo1@vt.edu
                elipsco1@vt.edu
                apruden@vt.edu
                heath@vt.edu
                pvikes@vt.edu
                lqzhang@cs.vt.edu
                Journal
                Microbiome
                Microbiome
                Microbiome
                BioMed Central (London )
                2049-2618
                1 February 2018
                1 February 2018
                2018
                : 6
                : 23
                Affiliations
                [1 ]ISNI 0000 0001 0694 4940, GRID grid.438526.e, Department of Computer Science, , Virginia Tech, ; Blacksburg, VA USA
                [2 ]ISNI 0000 0001 0694 4940, GRID grid.438526.e, Department of Civil and Environmental Engineering, , Virginia Tech, ; Blacksburg, VA USA
                Article
                401
                10.1186/s40168-018-0401-z
                5796597
                29391044
                a6ec8b9a-6e92-4e8b-b332-b5ca4a9f5c90
                © The Author(s). 2018

                Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

                History
                : 29 June 2017
                : 10 January 2018
                Funding
                Funded by: FundRef http://dx.doi.org/10.13039/100000199, U.S. Department of Agriculture;
                Award ID: 2015-68003-23050
                Award Recipient :
                Funded by: FundRef http://dx.doi.org/10.13039/100000001, National Science Foundation;
                Award ID: 1545756
                Award Recipient :
                Categories
                Software
                Custom metadata
                © The Author(s) 2018

                metagenomics,antibiotic resistance,deep learning,machine learning

                Comments

                Comment on this article