CoChat
Sign Up
Cited by 562Genomics and Phylogenetic StudiesScientific Computing and Data ManagementMicrobial Community Ecology and Physiology

MGnify: the microbiome analysis resource in 2020

Open access
Mitchell, Alex L, Almeida, Alexandre, Beracochea, Martin, Boland, Miguel, Burgin, Josephine, Cochrane, Guy, Crusoe, Michael R, Kale, Varsha +9 more·2019·Nucleic Acids Research
DOI 10.1093/nar/gkz1035PMC PMC7145632
Open this paper in CoChat — free accountOpen PDF

Abstract

MGnify (http://www.ebi.ac.uk/metagenomics) provides a free to use platform for the assembly, analysis and archiving of microbiome data derived from sequencing microbial populations that are present in particular environments. Over the past 2 years, MGnify (formerly EBI Metagenomics) has more than doubled the number of publicly available analysed datasets held within the resource. Recently, an updated approach to data analysis has been unveiled (version 5.0), replacing the previous single pipeline with multiple analysis pipelines that are tailored according to the input data, and that are formally described using the Common Workflow Language, enabling greater provenance, reusability, and reproducibility. MGnify's new analysis pipelines offer additional approaches for taxonomic assertions based on ribosomal internal transcribed spacer regions (ITS1/2) and expanded protein functional annotations. Biochemical pathways and systems predictions have also been added for assembled contigs. MGnify's growing focus on the assembly of metagenomic data has also seen the number of datasets it has assembled and analysed increase six-fold. The non-redundant protein database constructed from the proteins encoded by these assemblies now exceeds 1 billion sequences. Meanwhile, a newly developed contig viewer provides fine-grained visualisation of the assembled contigs and their enriched annotations.

AI snapshot

Written from the paper’s full text.

TL;DR

MGnify’s 2020 update reports major growth of its microbiome analysis resource and the release of version 5.0, which replaces a single pipeline with multiple data-type-specific pipelines formally described in Common Workflow Language. It adds ITS-based taxonomy, expanded functional annotations, assembly-derived pathway/system predictions, a contig viewer, and a greatly expanded non-redundant protein database.

Key findings

  • MGnify contains over 3,500 publicly available projects, comprising ∼175,000 samples, ∼230,000 runs and ∼240,000 analyses, a 2-fold increase in analysed datasets over two years; most analyses are 16S/18S rRNA amplicon (∼193,000), followed by metagenomic (∼27,000), assembly (∼16,000), metatranscriptomic (∼2,000) and metabarcoding (∼2,000).
  • Pipeline v5.0, released September 2019, adds ITS1/2 taxonomic classification, KEGG orthologue predictions for reads and assembled contigs, and KEGG pathway and Genome Properties annotations for assembled contigs.
  • MGnify has assembled ∼43,000 sequencing runs, mostly with metaSPAdes (<0.1% not due to memory limits), with about one third uploaded and analysed; the MGnify protein database grew from 50 million sequences (v2017_08) to over 1.1 billion (v2019_05), with only 9 million identical sequences shared with UniProtKB.
  • A new contig viewer uses the Integrative Genomics Viewer framework, and “Super Studies” group related projects, such as all 96 Earth Microbiome Project studies and 8 Tara Oceans datasets.

Methods

  • This is a resource-update paper describing MGnify’s data model, accessions (MGYS, MGYA, MGYP), ENA/INSDC integration, and analysis of user-submitted and public microbiome datasets.
  • Version 5.0 uses separate CWL-described pipelines for amplicon, metagenomic/metatranscriptomic raw reads, and assemblies, executed with toil-cwl-runner; tools include SeqPrep, Trimmomatic, Rfam, MAPseq with SILVA/ITSoneDB/UNITE, mOTUs2, FragGeneScan, InterPro, Pfam, KEGG/KOfam, Prodigal, eggNOG-mapper, antiSMASH, and DIAMOND against UniRef90.
  • The protein database was clustered with Linclust from MMseqs using 90% thresholds for overlap and sequence identity, and online searches are provided against cluster representatives.

Limitations

  • Stated: private assembled datasets cannot be brokered into ENA by MGnify, so users must submit assemblies under their own credentials, adding burden; containerization of all pipeline tools for cloud deployment remains ongoing; online sequence similarity searches against the entire protein dataset are not possible due to size, so only cluster representatives are searchable.
  • Stated: assembly is computationally expensive and can require terabytes of memory with metaSPAdes; a handful of datasets (<0.1%) could not use metaSPAdes because memory requirements exceeded available maximum.
  • Evident from the text: this is a resource and pipeline update rather than a benchmark or validation study, and it does not report accuracy comparisons or independent performance evaluations of the new v5.0 analyses.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1093/nar/gkz1035

Cites

53 of 53 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 16 cited sources checked · 0 supported · 0 flagged

  • Findingnone given

    MGnify is a free-to-use platform for the assembly, analysis and archiving of microbiome data derived from sequencing microbial populations present in particular environments.

    “MGnify (http://www.ebi.ac.uk/metagenomics) provides a free to use platform for the assembly, analysis and archiving of microbiome data derived from sequencing microbial populations that are present in particular environments.”
    • Also stated in: § INTRODUCTION · plain
      “MGnify (previously known as EBI Metagenomics (14)) is a freely available hub for the analysis, exploration and archiving of microbiome data.”
    Quote in the paperNot a substantial claim
  • Findingown data

    Over the past two years, MGnify has more than doubled the number of publicly available analysed datasets held within the resource.

    “Over the past 2 years, MGnify (formerly EBI Metagenomics) has more than doubled the number of publicly available analysed datasets held within the resource.”
    • Also stated in: § UPDATES TO MGnify CONTENT · plain
      “This represents a 2-fold increase in the number of datasets analysed over the last two years.”
    Quote in the paperNot a substantial claim
  • Findingnone given

    Biochemical pathways and systems predictions have been added for assembled contigs.

    “Biochemical pathways and systems predictions have also been added for assembled contigs.”
    Quote in the paperNot a substantial claim
  • Findingnone given

    At the renaming of MGnify in June 2018, unique and stable accession numbers for studies and analyses were introduced, replacing study and run accessions inherited from INSDC.

    “At the time of the renaming of resource in June 2018, unique and stable accession numbers for studies and analyses in MGnify were introduced, replacing the study and run accessions, respectively, that were inherited from INSDC.”
    Quote in the paperNot a substantial claim
  • Findingnone given

    MGnify version 5.0 replaced the previous single pipeline with multiple analysis pipelines tailored according to input data.

    “Recently, an updated approach to data analysis has been unveiled (version 5.0), replacing the previous single pipeline with multiple analysis pipelines that are tailored according to the input data, and that are formally described using the Common Workflow Language, enabling greater provenance, reusability, and reproducibility.”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim. 22
  • May combine two claims. 19

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    MGnify covers studies targeting taxonomic markers such as SSU rRNA gene (amplicon), whole genome shotgun sequencing (metagenomics), and whole transcriptome shotgun sequencing (metatranscriptomics).

    “Amongst the data types covered are studies that target taxonomic markers, such as the small subunit (SSU) ribosomal ribonucleic acid gene (amplicon studies), whole genome shotgun sequencing studies (metagenomics) and whole transcriptome shotgun sequencing studies (metatranscriptomics).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingcitation

    Through partnership with ENA, sequencing data and metadata submitted for analysis are accessioned and stored permanently within the archive, which operates under INSDC.

    “Through partnership with the ENA, sequencing data and metadata submitted for analysis are accessioned and stored permanently within the archive, which operates under the International Nucleotide Sequence Database Collaboration (INSDC) (15).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [15] Karsch-Mizrachi, 2018, The international nucleotide sequence database collaboration, Nucleic Acids Res. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    In MGnify, analysis of user-submitted data is supplemented by processing publicly available microbiome datasets drawn from INSDC via ENA.

    “In MGnify, analysis of user-submitted data is supplemented by processing publicly available microbiome datasets drawn from the INSDC via the ENA.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Findingnone given

    MGnify accepts user-submitted data and provides standardized pipelines that offer taxonomic and, where appropriate, functional analysis of microbiome datasets.

    “The resource accepts user-submitted data and provides standardized pipelines that offer taxonomic and (where appropriate) functional analysis of microbiome datasets.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    MGnify curators add supplementary metadata to imported studies, assigning a relevant biome selected from GOLD biosample hierarchy.

    “The distinction between MGnify studies and ENA projects is important, as MGnify curators add supplementary metadata to imported studies, assigning a relevant biome selected from the Genomes OnLine Database (GOLD) (17) biosample hierarchy.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [17] Mukherjee, 2019, Genomes OnLine database (GOLD) v.7: updates and new features, Nucleic Acids Res. source

      Not checked in the free run.

  • May combine two claims.

    Findingnone given

    MGnify's new analysis pipelines offer additional approaches for taxonomic assertions based on ribosomal internal transcribed spacer regions (ITS1/2) and expanded protein functional annotations.

    “MGnify's new analysis pipelines offer additional approaches for taxonomic assertions based on ribosomal internal transcribed spacer regions (ITS1/2) and expanded protein functional annotations.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    MGnify analyses reside within MGnify, which provides an independent website and API for data discovery and exploration.

    “Meanwhile, the analyses of these datasets (both pre-release and publicly available) reside within MGnify, which provides an independent website and API for data discovery and exploration.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    MGnify users can request that any relevant publicly available study found within INSDC can be analysed with the latest MGnify pipeline version, and the results added to the resource.

    “Furthermore, MGnify users can request that any relevant publicly available study found within the INSDC can be analysed with the latest MGnify pipeline version, and the results added to the resource.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    MGnify has begun to offer analysis of user-submitted assembled sequence data and/or can provide assembly of user-submitted metagenomic data prior to analysis (available upon request).

    “More recently, the resource has begun to offer analysis of user-submitted assembled sequence data (assembly), and/or can provide assembly of user-submitted metagenomic data prior to analysis (available upon request).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    To date, MGnify has assembled approximately 43000 sequencing runs drawn from a wide range of biomes, and is in the process of submitting the assembled contigs back into ENA.

    “To date, we have assembled ∼43 000 sequencing runs drawn from a wide range of biomes, and are in the process of submitting the assembled contigs back into the ENA.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    Since the first inception of the analysis pipeline in 2011, there have been five major versions.

    “Since the first inception of the analysis pipeline in 2011, there have been five major versions.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Additional functional analyses are provided in the form of KEGG orthologue annotations, calculated using HMMER and the KOfam library on predicted protein sequences.

    “Additional functional analyses are also provided in the form of KEGG orthologue (KO) (23) annotations, which are calculated using HMMER (40) and the KOfam library (using a slightly modified form of the profile hidden Markov model (HMM) database of KEGG orthologues (41)) on the predicted protein sequences.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [23] Kanehisa, 2016, KEGG as a reference resource for gene and protein annotation, Nucleic Acids Res. source

      Not checked in the free run.

    • cites [40] Eddy, 2011, Accelerated Profile HMM Searches, PLoS Comput. Biol. source

      Not checked in the free run.

    • cites [41] Aramaki, 2019, KofamKOALA: KEGG ortholog assignment based on profile HMM and adaptive score threshold source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Only 9 million identical sequences are in common between MGnify and UniProtKB, indicating very little overlap.

    “Despite the substantial increase in number, there remains very little overlap with the sequences found in MGnify and those in UniProtKB, with only 9 million identical sequences in common between the two resources.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingnone given

    The different proportions of analyses (amplicon, metagenomics and assembled) closely reflect the corresponding proportions of raw sequence data found in ENA.

    “The different proportions of analyses (amplicon, metagenomics and assembled) closely reflect the corresponding proportions of raw sequence data found in ENA.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingcitation

    MGnify's latest update v5.0, released September 2019, extends analyses further, substantially increasing the scope of functional annotations through the addition of KEGG orthologue predictions for reads and assembled contigs, and KEGG pathways and Genome Properties annotations for assembled contigs.

    “The latest update (v5.0), released September 2019, extends the analyses further, substantially increasing the scope of functional annotations through the addition of KEGG (23) orthologue predictions for reads and assembled contigs, and KEGG pathways and Genome Properties (24) annotations for assembled contigs.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [23] Kanehisa, 2016, KEGG as a reference resource for gene and protein annotation, Nucleic Acids Res. source

      Not checked in the free run.

    • cites [24] Richardson, 2019, Genome properties in 2019: a new companion database to InterPro for the inference of complete functional attributes, Nucleic Acids Res. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Approximately 25% of MGnify protein sequences are predicted to be full length according to Prodigal's gene identification criteria.

    “Approximately 25% of these sequences are predicted to be full length according to Prodigal's gene identification criteria, with 15% truncated at either the N or C terminus and 60% predicted to be truncated at both ends (termed ‘partials’).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The majority of MGnify data are 16S or 18S rRNA gene amplicon datasets, followed by metagenomic and assembly analyses, with smaller numbers of metatranscriptomes and metabarcoding runs.

    “The majority of this data (∼193 000 analyses) are 16S or 18S rRNA gene amplicon datasets, followed by metagenomic (∼27 000) and assembly (∼16 000) analyses, with a smaller number of analysed metatranscriptomes (∼2000) and metabarcoding runs (∼2000, the majority of which target internally transcribed spacer (ITS) regions - as previous pipelines could not analyse ITS regions, these datasets will be”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning

    Due to the clustering parameters, the selected cluster representative (centroid sequence) is typically the longest.

    “Due to the clustering parameters, the selected cluster representative (the centroid sequence), is typically the longest.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    At present, all 96 individual studies that make up the Earth Microbiome Project are grouped together under a single Super Study, as are MGnify analyses for 8 distinct Tara Oceans datasets.

    “For example, at present, all 96 individual studies that make up the Earth Microbiome Project (18) are grouped together under a single Super Study, as are MGnify analyses for 8 distinct Tara Oceans (19) datasets.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [18] Thompson, 2017, A communal catalogue reveals Earth's multiscale microbial diversity, Nature source

      Not checked in the free run.

    • cites [19] Karsenti, 2011, A holistic approach to marine eco-systems biology, PLoS Biol. source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    MGnify currently contains over 3500 publicly available projects, comprising approximately 175000 samples, 230000 runs and 240000 analyses.

    “In terms of overall data, MGnify currently contains over 3500 publicly available projects, comprising ∼175 000 samples, ∼230 000 runs and ∼240 000 analyses.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    In version 5.0, analysis of amplicon datasets has been expanded from SSU and LSU-based classifications to include those based on ITS regions.

    “In version 5.0 of the pipeline, analysis of amplicon datasets has been expanded from SSU and LSU-based classifications to include those based on ITS regions.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingcitation

    The new MGnify pipeline defined the CWL workflow first and utilizes toil-cwl-runner to orchestrate execution on their compute infrastructure.

    “The new pipeline is a major departure from this practice, where we have defined the CWL workflow first, and now utilize toil-cwl-runner (a CWL compatibility layer that is a part of the Toil workflow execution engine (26)) to orchestrate the execution on our compute infrastructure.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [26] Vivian, 2017, Toil enables reproducible, open source, big biomedical data analyses, Nat. Biotechnol. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning

    Pfam annotations are now provided as an additional results set to aid comparison with other resources that use Pfam for annotation, such as IMG/M.

    “While previously available as part of the InterPro results, Pfam (39) annotations are now provided as an additional results set to aid comparison with other resources that use Pfam for annotation, such as IMG/M.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [39] El-Gebali, 2019, The Pfam protein families database in 2019, Nucleic Acids Res. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning

    The two ITS reference databases (ITSoneDB and UNITE) offer complementary coverage.

    “The two ITS reference databases offer complementary coverage.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    Enabling large scale data analysis using standardized pipelines allows studies to be placed in context with each other, increasing data reuse and maximizing knowledge extracted from datasets.

    “Enabling large scale data analysis using standardized pipelines in this way allows studies to be placed in context with each other, increasing data reuse and maximizing the knowledge that can be extracted from the datasets.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The introduction of unique stable accession numbers allows users to distinguish analysed microbiome datasets from primary sequence data in ENA and helps identify where MGnify adds content.

    “Driven by ongoing efforts to adhere to the FAIR data principles (16) (aimed to make data findable, accessible, interoperable and reusable), this change allows users to distinguish analysed microbiome datasets from the primary sequence data in ENA - which continue to be presented under the study and run accessions - and helps identify where MGnify adds content.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [16] Wilkinson, 2016, The FAIR Guiding Principles for scientific data management and stewardship, Sci. Data source

      Not checked in the free run.

  • May combine two claims.

    Study design and conductreasoning

    Defining the CWL workflow first and using toil-cwl-runner removed duplication of effort and helps ensure provenance, interoperability and reproducibility of the pipeline.

    “This has removed the duplication of effort (in terms of implementing the logic in software and then retrospectively producing a CWL description), and helps ensure provenance, interoperability and reproducibility of the pipeline.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    The adoption of third party workflow execution engines reduces the burden for maintaining pipelining software.

    “The adoption of third party workflow execution engines reduces the burden for maintaining pipelining software”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    KEGG pathways and Genome Properties annotations represent higher order annotations (based on two or more proteins) that enable the description of functional systems encoded by the microbes within a metagenome.

    “The latter two represent higher order annotations (based on two or more proteins) that enable the description of functional systems encoded by the microbes within a metagenome.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductcitation

    Additional phylogenetic marker gene-based operational taxonomic unit profiling is performed using mOTUs2 on all reads that pass QC, which allows sensitive and accurate quantification for both known and unknown species.

    “Additional phylogenetic marker gene-based operational taxonomic unit profiling is performed using mOTUs2 (35) on all reads that pass QC, which allows sensitive and accurate quantification for both known and unknown species.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [35] Milanese, 2019, Microbial abundance, activity and population genomic profiling with mOTUs2, Nat. Commun. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Metagenomics has provided the scientific community with access to approximately 99% of organisms that are yet to be cultured.

    “Metagenomics has started to provide the scientific community with access to the ∼99% of organisms that are yet to be cultured (50).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [50] Rinke, 2013, Insights into the phylogeny and coding potential of microbial dark matter, Nature source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    The Super Studies facility makes it easier to discover, explore and compare data across large-scale projects.

    “The Super Studies facility makes it easier to discover, explore and compare data across such large-scale projects.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Interpretationnone given

    Assembly of reads into longer contigs can underpin a deeper understanding of the data through recovery of full-length proteins, prediction of complete biochemical pathways and/or biological systems, and potential reconstruction of complete genomes.

    “Whilst analysis of raw reads can provide detailed insights into the microbiome, assembly of the reads into longer contigs can underpin a deeper understanding of the data through recovery of full-length proteins, prediction of complete biochemical pathways and/or biological systems, and potential reconstruction of complete genomes.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductcitation

    metaSPAdes is MGnify's primary assembly algorithm because the ensuing assemblies are typically of better quality than those produced using other assemblers.

    “Assembly is a computationally expensive process, frequently requiring terabytes of memory to accomplish when using metaSPAdes (20), which is our primary assembly algorithm, as the ensuing assemblies are typically of better quality than those produced using other assemblers (21).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [20] Nurk, 2017, metaSPAdes: a new versatile metagenomic assembler, Genome Res. source

      Not checked in the free run.

    • cites [21] Vollmers, 2017, Comparing and evaluating metagenome assembly tools from a microbiologist's perspective - not only size matters!, PLoS One source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    These updates have particularly improved the interpretation of metagenomics assemblies.

    “Notably, these have particularly improved the interpretation of metagenomics assemblies.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The assembled data provides a deeper understanding of microbiomes, providing access to full length proteins, enabling the assertion of pathways and systems, and is foundational to the reconstruction of genomes.
35 of 35

  • The paper relies on this without support.

    MGnify is one of the world's largest resources for the analysis of microbiome datasets.

    “MGnify is one of the world's largest resources for the analysis of microbiome datasets.”

  • The paper relies on this without support.

    MGnify and IMG/M are the only major resources offering metagenomic assembly as a service.

    “To our knowledge, MGnify and IMG/M are the only major resources offering metagenomic assembly as a service, with both having large, inequivalent collections of metagenomics assemblies.”

  • The paper relies on this without support.

    The assembled data provides a deeper understanding of microbiomes, providing access to full length proteins, enabling the assertion of pathways and systems, and is foundational to the reconstruction of genomes.

    “While assembly of large shotgun metagenomics datasets remains computationally expensive, the assembled data (potentially) provides a deeper understanding of microbiomes, providing access to full length proteins, enabling the assertion of pathways and systems, and are foundational to the reconstruction of genomes (i.e. metagenome assembled genomes, known as ‘MAGs’).”

  • The paper relies on this without support.

    Understanding the genomic context of proteins can provide greater insights into their functional roles.

    “However, understanding the genomic context of proteins can provide greater insights into their functional roles.”

  • The paper relies on this without support.

    A metagenomics assembly may typically contain more than 1000 contigs.

    “As a metagenomics assembly may typically contain >1000 contigs, a range of parameters can be selected to filter the results.”
5 of 5

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The paper folds over 617 million sequences from the MGnify90 database.

    “We fold over 617 million sequences from the MGnify90 database (25).”

    From Evolutionary-scale prediction of atomic level protein structure with a language model (2022) · cites this paper as [25]

    The 'UPDATES TO THE MGnify PROTEIN DATABASE' section describes MGnify protein sequences clustered at 90% identity, supporting the identification of the database. It reports over 1.1 billion sequences in the full database, but does not give a count of 617 million for the 90%-clustered set. The citing authors' folding of those sequences is their own work, not a finding attributed to this source. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

  • The ESM Metagenomic Atlas is the first large-scale structural characterization of metagenomic proteins, with more than 617 million structures, folding practically all sequences in MGnify90.

    “This is the first large-scale structural characterization of metagenomic proteins, with more than 617 million structures.”

    From Evolutionary-scale prediction of atomic level protein structure with a language model (2022) · cites this paper as [25]

    The ‘UPDATES TO THE MGnify PROTEIN DATABASE’ section describes more than 1.1 billion protein sequences and their clustering, not predicted structures. Searches for the ESM Metagenomic Atlas, 617 million structures, and protein folding found no passage addressing the claimed structural characterization. (LLM, conf=0.76)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • SeqPrep citation points to an AdapterRemoval paper

    Minor

    The amplicon-pipeline description cites reference 27 for SeqPrep, but reference 27 is a paper about AdapterRemoval.

1 of 1

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Mitchell, A. L., Almeida, A., Beracochea, M., Boland, M., Burgin, J., Cochrane, G., Crusoe, M. R., Kale, V., Potter, S. C., Richardson, L. J., Sakharova, E., Scheremetjew, M., Korobeynikov, A., Shlemov, A., Kunyavskaya, O., Lapidus, A., & Finn, R. D.
Affiliations
  • European Bioinformatics Institute (GB)
  • European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
  • Common Workflow Language, a project of the Software Freedom Conservancy, Inc. 137 Montague Street, Suite 380, Brooklyn, NY 11201-3548, USA
  • St Petersburg University (RU)
  • Center for Algorithmic Biotechnologies, Saint Petersburg State University, Russia
Corresponding author
Finn, Robert D
ORCID
Venue
Nucleic Acids Research
Publisher
Oxford University Press (OUP)
Published
2019-11-07
Volume
48
Issue
D1
Pages
D570-D578
ISSN
  • 0305-1048 (ISSN-L)
  • 1362-4962
DOI
10.1093/nar/gkz1035
PMC
PMC7145632
OpenAlex
W2984894304
Semantic Scholar
633155ae06b2367cc42d5beb21d2be85330b6200
Cited by
562
References
53
Citations by year
2026: 23 · 2025: 68 · 2024: 100 · 2023: 112 · 2022: 131 · 2021: 95 · 2020: 30 · 2019: 2
FWCI
21.33
Type
journal article
Language
en
Open access
Licence
Locations
  • Nucleic Acids Research (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • Horizon 2020 (676559, 817729)
  • Biotechnology and Biosciences Research Council (BB/M011755/1, BB/R015228/1, BB/N018354/1)
  • ELIXIR
  • Russian Fund for Basic Research (18-54-74004)
  • European Molecular Biology Laboratory
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
  • Social Sciences › Decision Sciences › Information Systems and Management › Scientific Computing and Data Management
  • Physical Sciences › Environmental Science › Ecology › Microbial Community Ecology and Physiology
Keywords
MGnify, metagenome assembly, data analysis pipeline, Common Workflow Language, reproducibility, taxonomic annotation, ITS1-2, protein annotation, microbial communities
MeSH terms
Archaea, Bacteria, Phylogeny, Software, DNA, Ribosomal Spacer, Databases, Genetic, Metagenome, Metagenomics, Microbiota

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-10.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-10.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review