CoChat

MetaEuk—sensitive, high-throughput gene discovery, and annotation for large-scale eukaryotic metagenomics

Open access
Levy Karin, Eli, Mirdita, Milot, Söding, Johannes · 2020 · Microbiome

Published version of a preprint

The DOI matches its Crossref record; 66 of the 66 references it cites were found in a registry or an index.

Abstract

Background: Metagenomics is revolutionizing the study of microorganisms and their involvement in biological, biomedical, and geochemical processes, allowing us to investigate by direct sequencing a tremendous diversity of organisms without the need for prior cultivation. Unicellular eukaryotes play essential roles in most microbial communities as chief predators, decomposers, phototrophs, bacterial hosts, symbionts, and parasites to plants and animals. Investigating their roles is therefore of great interest to ecology, biotechnology, human health, and evolution. However, the generally lower sequencing coverage, their more complex gene and genome architectures, and a lack of eukaryote-specific experimental and computational procedures have kept them on the sidelines of metagenomics. Results: MetaEuk is a toolkit for high-throughput, reference-based discovery, and annotation of protein-coding genes in eukaryotic metagenomic contigs. It performs fast searches with 6-frame-translated fragments covering all possible exons and optimally combines matches into multi-exon proteins. We used a benchmark of seven diverse, annotated genomes to show that MetaEuk is highly sensitive even under conditions of low sequence similarity to the reference database. To demonstrate MetaEuk’s power to discover novel eukaryotic proteins in large-scale metagenomic data, we assembled contigs from 912 samples of the Tara Oceans project. MetaEuk predicted >12,000,000 protein-coding genes in 8 days on ten 16-core servers. Most of the discovered proteins are highly diverged from known proteins and originate from very sparsely sampled eukaryotic supergroups. Conclusion: The open-source (GPLv3) MetaEuk software ( https://github.com/soedinglab/metaeuk ) enables large-scale eukaryotic metagenomics through reference-based, sensitive taxonomic and functional annotation.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/s40168-020-00808-x

Cites

66 of 66 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 MetaEuk algorithm. Input to MetaEuk are assembled metagenomic contigs and a reference database of protein sequences. 1 Six-frame translation of all putative protein-coding fragments from each contig. 2 Fragments on the same contig and strand that hit the same reference protein T are examined together. 3 Putative exons are identified and ordered according to their start position on the contig. The highest score and path (denoted with a star) of a set of compatible putative exons is computed by dynamic programming, in which individual scores of the putative exons are summed and unmatched amino acids are penalized. 4 Redundancies among gene calls due to homologous targets (T, T’, and “T”) are reduced and a representative prediction (denoted with a star) is retained. 5 Contradicting predictions of overlapping genes on the same strand are resolved by excluding the prediction with the higher E-value
Fig. 1 MetaEuk algorithm. Input to MetaEuk are assembled metagenomic contigs and a reference database of protein sequences. 1 Six-frame translation of all putative protein-coding fragments from each contig. 2 Fragments on the same contig and strand that hit the same reference protein T are examined together. 3 Putative exons are identified and ordered according to their start position on the contig. The highest score and path (denoted with a star) of a set of compatible putative exons is computed by dynamic programming, in which individual scores of the putative exons are summed and unmatched amino acids are penalized. 4 Redundancies among gene calls due to homologous targets (T, T’, and “T”) are reduced and a representative prediction (denoted with a star) is retained. 5 Contradicting predictions of overlapping genes on the same strand are resolved by excluding the prediction with the higher E-valueCC BY · PubMed Central
Fig. 2 MetaEuk evaluation on benchmark. MetaEuk predictions were mapped to annotated proteins. a Conditions of increasing evolutionary divergence were simulated by excluding gene calls based on their sequence identity to their target. Sensitivity is the fraction of annotated proteins from the query genome to which a MetaEuk prediction was mapped. b Fraction of exons covered by MetaEuk (color saturation). The number of MetaEuk predictions is indicated on top of each bar. c In an annotation-dependent precision estimation, MetaEuk predictions that mapped to an annotated protein were considered as “true” and the rest as “false.” These sets of predictions are well separated by their E-values, as indicated by the high AUC-PR values. d Fraction of annotated protein-coding genes that were split by MetaEuk into two (dark grey) or three (black) different predictions. e Comparison of the E-values computed by MetaEuk and by the Smith-Waterman algorithm for A. castellani proteins. The Spearman rho values indicate high correlation for A. castellani and for the other organisms (Supp. Figure S3A)
Fig. 2 MetaEuk evaluation on benchmark. MetaEuk predictions were mapped to annotated proteins. a Conditions of increasing evolutionary divergence were simulated by excluding gene calls based on their sequence identity to their target. Sensitivity is the fraction of annotated proteins from the query genome to which a MetaEuk prediction was mapped. b Fraction of exons covered by MetaEuk (color saturation). The number of MetaEuk predictions is indicated on top of each bar. c In an annotation-dependent precision estimation, MetaEuk predictions that mapped to an annotated protein were considered as “true” and the rest as “false.” These sets of predictions are well separated by their E-values, as indicated by the high AUC-PR values. d Fraction of annotated protein-coding genes that were split by MetaEuk into two (dark grey) or three (black) different predictions. e Comparison of the E-values computed by MetaEuk and by the Smith-Waterman algorithm for A. castellani proteins. The Spearman rho values indicate high correlation for A. castellani and for the other organisms (Supp. Figure S3A)CC BY · PubMed Central
Fig. 3 Reference profiles composition. Proteins from three datasets: MERC (292 million), MMETSP (18.5 million), and Uniclust50 (21 million) were clustered into ~ 88 million clusters. Most clusters contained proteins from a single reference dataset. The profiles computed based on these clusters served as the reference database for the MetaEuk run on the Tara Oceans contigs
Fig. 3 Reference profiles composition. Proteins from three datasets: MERC (292 million), MMETSP (18.5 million), and Uniclust50 (21 million) were clustered into ~ 88 million clusters. Most clusters contained proteins from a single reference dataset. The profiles computed based on these clusters served as the reference database for the MetaEuk run on the Tara Oceans contigsCC BY · PubMed Central
Fig. 4 MetaEuk predictions on Tara Oceans contigs. MetaEuk was run on over 1.3 million contigs assembled from Tara Oceans metagenomic reads against a reference database of ~88 million protein profiles. a The number of MetaEuk predictions per contig increases with its length. Horizontal lines mark contig length quartiles. b Most (76%) MetaEuk predictions had a single putative exon. The absolute number of predictions is indicated above each bar. c Single-exon predictions tend to have longer putative exons than multi-exon predictions
Fig. 4 MetaEuk predictions on Tara Oceans contigs. MetaEuk was run on over 1.3 million contigs assembled from Tara Oceans metagenomic reads against a reference database of ~88 million protein profiles. a The number of MetaEuk predictions per contig increases with its length. Horizontal lines mark contig length quartiles. b Most (76%) MetaEuk predictions had a single putative exon. The absolute number of predictions is indicated above each bar. c Single-exon predictions tend to have longer putative exons than multi-exon predictionsCC BY · PubMed Central
Fig. 5 MetaEuk predictions compared to the reference datasets. MetaEuk predicted proteins were queried against the representative sequence of their target reference cluster. a About one third of the predicted MetaEuk proteins had less than 60% sequence identity to their target. b Targets are well covered by MetaEuk predicted proteins
Fig. 5 MetaEuk predictions compared to the reference datasets. MetaEuk predicted proteins were queried against the representative sequence of their target reference cluster. a About one third of the predicted MetaEuk proteins had less than 60% sequence identity to their target. b Targets are well covered by MetaEuk predicted proteinsCC BY · PubMed Central
Fig. 6 Taxonomy of Tara Oceans contigs with MetaEuk predictions. The best-scoring taxonomic label of all predictions on each contig was conferred to the contig. Contigs were divided into four categories according to their number of MetaEuk predictions. Over 82% of the contigs were assigned to the domain Eukaryota. a The proportion of unassigned contigs decreases with the number of MetaEuk predictions on the contig. The fraction of eukaryotic contigs out of all assigned contigs is about 90% in all four categories. b Eukaryotic taxonomic labels below the domain level
Fig. 6 Taxonomy of Tara Oceans contigs with MetaEuk predictions. The best-scoring taxonomic label of all predictions on each contig was conferred to the contig. Contigs were divided into four categories according to their number of MetaEuk predictions. Over 82% of the contigs were assigned to the domain Eukaryota. a The proportion of unassigned contigs decreases with the number of MetaEuk predictions on the contig. The fraction of eukaryotic contigs out of all assigned contigs is about 90% in all four categories. b Eukaryotic taxonomic labels below the domain levelCC BY · PubMed Central
Fig. 7 Diversity of MetaEuk marine eukaryotic proteins. Homologous sequences of the large subunits of RNA polymerases of 985 species as well as 1076 MetaEuk marine proteins were collected and a maximum-likelihood tree was computed based on their alignment. MetaEuk sequences (black) expand major eukaryotic lineages, including deeply rooted supergroups (denoted with star)
Fig. 7 Diversity of MetaEuk marine eukaryotic proteins. Homologous sequences of the large subunits of RNA polymerases of 985 species as well as 1076 MetaEuk marine proteins were collected and a maximum-likelihood tree was computed based on their alignment. MetaEuk sequences (black) expand major eukaryotic lineages, including deeply rooted supergroups (denoted with star)CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

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

The self-check needs the paper’s full text.

Citations

Cites

Show all 66 references

Cited by

See all 316

Metadata

Authors
Levy Karin, E., Mirdita, M., & Söding, J.
Affiliations
Max Planck Institute for Biophysical Chemistry (DE)
Corresponding author
Söding, Johannes
Venue
Microbiome
Published
2020-04-03
Volume
8
Issue
1
Pages
48-48
ISSN
2049-2618 (ISSN-L)
OpenAlex
W3014607123
Cited by
299
References
65
Citations by year
2026: 51 · 2025: 64 · 2024: 72 · 2023: 52 · 2022: 33 · 2021: 21 · 2020: 5
FWCI
8.96
Type
journal article
Language
en
Open access
Licence
Locations
  • Microbiome (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • GoeScholar The Publication Server of the Georg-August-Universität Göttingen (Georg-August-Universität Göttingen) (submittedVersion, cc-by)
  • MPG.PuRe (Max Planck Society) (submittedVersion, other-oa)
Funders
Horizon 2020 Framework Programme (685778)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
  • Physical Sciences › Environmental Science › Ecology › Environmental DNA in Biodiversity Studies
  • Life Sciences › Agricultural and Biological Sciences › Ecology, Evolution, Behavior and Systematics › Protist diversity and phylogeny
Keywords
genome annotation, Tara Oceans, eukaryotic supergroups, taxonomic annotation, functional annotation, metagenomic contigs
MeSH terms
Algorithms, Sequence Analysis, DNA, Computational Biology, Databases, Genetic, Metagenome, Metagenomics, Eukaryota, High-Throughput Screening Assays, Molecular Sequence Annotation, 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.

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: MetaEuk—sensitive, high-throughput gene discovery, and annotation for large-scale eukaryotic metagenomics. https://app.cochat.ai/papers/10.1186/s40168-020-00808-x
The paper
Levy Karin, E., Mirdita, M., & Söding, J. (2020). MetaEuk—sensitive, high-throughput gene discovery, and annotation for large-scale eukaryotic metagenomics. Microbiome. https://doi.org/10.1186/s40168-020-00808-x

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.