Metagenomics is the study of genetic material recovered directly from environmental samples, such as soil, water, or gut contents, without the need for isolating or cultivating individual organisms. Metagenomics binning is a process used to classify DNA sequences obtained from metagenomic sequencing into discrete groups, or bins, based on their similarity to each other.

The goal of metagenomics binning is to assign DNA sequences to the organisms or taxonomic groups from which they originate, allowing for a better understanding of the diversity and functions of the microbial communities present in the sample. This is typically achieved through computational methods that include sequence similarity, composition, and other features to group the sequences into bins.

Comment

Before starting this tutorial, it is recommended to do the Metagenomics Assembly Tutorial

Binning approaches

There are several approaches to metagenomics binning, including:

• Sequence composition-based binning: This method is based on the observation that different genomes have distinct sequence composition patterns, such as GC content or codon usage bias. By analyzing these patterns in metagenomic data, sequence fragments can be assigned to individual genomes or groups of genomes.

• Coverage-based binning: This method uses the depth of coverage of sequencing reads to group them into bins. Sequencing reads that originate from the same genome are expected to have similar coverage, and this information can be used to identify groups of reads that represent individual genomes or genome clusters.

• Hybrid binning: This method combines sequence composition-based and coverage-based binning to increase the accuracy of binning results. By using multiple sources of information, hybrid binning can better distinguish closely related genomes that may have similar sequence composition patterns.

• Clustering-based binning: This method groups sequence fragments into clusters based on sequence similarity, and then assigns each cluster to a genome or genome cluster based on its sequence composition and coverage. This method is particularly useful for metagenomic data sets with high levels of genomic diversity.

• Supervised machine learning-based binning: This method uses machine learning algorithms trained on annotated reference genomes to classify metagenomic data into bins. This approach can achieve high accuracy but requires a large number of annotated genomes for training.

Binning challenges

Metagenomic binning is a complex process that involves many steps and can be challenging due to several problems that can occur during the process. Some of the most common problems encountered in metagenomic binning include:

• High complexity: Metagenomic samples contain DNA from multiple organisms, which can lead to high complexity in the data.
• Fragmented sequences: Metagenomic sequencing often generates fragmented sequences, which can make it difficult to assign reads to the correct bin.
• Uneven coverage: Some organisms in a metagenomic sample may be more abundant than others, leading to uneven coverage of different genomes.
• Incomplete or partial genomes: Metagenomic sequencing may not capture the entire genome of a given organism, which can make it difficult to accurately bin sequences from that organism.
• Horizontal gene transfer: Horizontal gene transfer (HGT) can complicate metagenomic binning, as it can introduce genetic material from one organism into another.
• Chimeric sequences: Sequences that are the result of sequencing errors or contamination can lead to chimeric sequences, which can make it difficult to accurately bin reads.
• Strain variation: Organisms within a species can exhibit significant genetic variation, which can make it difficult to distinguish between different strains in a metagenomic sample.

Common binners

There are different algorithms and tools performing metagenomic binning. The most widely used include:

• MaxBin (Wu et al. 2015): A popular de novo binning algorithm that uses a combination of sequence features and marker genes to cluster contigs into genome bins.
• MetaBAT (Kang et al. 2019): Another widely used de novo binning algorithm that employs a hierarchical clustering approach based on tetranucleotide frequency and coverage information.
• CONCOCT (Alneberg et al. 2014): A de novo binning tool that uses a clustering algorithm based on sequence composition and coverage information to group contigs into genome bins.
• MyCC (Lin and Liao 2016): A reference-based binning tool that uses sequence alignment to identify contigs belonging to the same genome or taxonomic group.
• GroopM (Imelfort et al. 2014): A hybrid binning tool that combines reference-based and de novo approaches to achieve high binning accuracy.
• SemiBin (Pan et al. 2022): A command-line tool for metagenomic binning with deep learning, handling both short and long reads.
• Vamb (Nissen et al. 2021): An algorithm that uses variational autoencoders (VAEs) to encode sequence composition and coverage information.
• ComeBin (Wang et al. 2024): A metagenomic binning tool that integrates both composition and abundance features with machine learning-based clustering to improve binning accuracy across complex microbial communities.

So many options, what binner to use ?

Each of these binning methods has its own strengths and limitations, and the choice of a binning tool often depends on the characteristics of the metagenomic dataset and the research question. Practical guidance on which binner to use for specific datasets and environments can be drawn from benchmark studies such as Author et al. 2025.

Open image in new tab

Another useful approach to investigate the performance of binners is to use simulated datasets. Therefore, the CAMI challenges (Meyer et al. 2022) were introduced. CAMI, which stands for Critical Assessment of Metagenome Interpretation, is an international community-driven initiative that organizes benchmarking challenges to objectively evaluate the performance of metagenomic tools. This includes the benchmarking of binners based on standardized, realistically simulated microbial communities that vary in complexity, strain diversity, and abundance.

Open image in new tab

A general approach is to perform binning using multiple binners that have shown good performance for the specific dataset, followed by bin refinement to generate an improved bin set that retains the best bins from the analysis.

Does using more binners always improve results? In practice, one must also consider computational resources and time constraints. Running many binners can be very time-consuming and resource-intensive, especially for large studies. In some cases, adding extra binners does not lead to a meaningful increase in bin quality, so the choice of binners should be made carefully. Overall, identifying the optimal combination of binners remains an active area of research, and clear, widely accepted guidelines are still being established.

Bin refinement

There are also bin refinement tools, which can evaluate, combine, and improve the raw bins produced by primary binners such as MetaBAT2, CONCOCT, MaxBin2, or SemiBin. These tools help remove contamination, merge complementary bins, and recover higher-quality MAGs.

• MetaWRAP (Uritskiy et al. 2018): A comprehensive metagenomic analysis pipeline that includes modules for quality control, assembly, binning (wrapping multiple binners), refinement, reassembly, and annotation. It provides an easy-to-use framework for producing high-quality MAGs from raw reads.

• DAS Tool (Sieber et al. 2018): A bin-refinement tool that combines results from multiple binners (e.g., MetaBAT2, MaxBin2, CONCOCT, SemiBin) into a consensus set of optimized, non-redundant bins. DAS Tool improves overall bin quality by integrating the strengths from several algorithms.

• Binnette (Mainguy and Hoede 2024): A fast and accurate bin refinement tool that constructs high-quality MAGs from the outputs of multiple binning tools. It generates hybrid bins using set operations on overlapping contigs — intersection, difference, and union — and evaluates their quality with CheckM2 to select the best bins. Compared to metaWRAP, Binette is faster and can process an unlimited number of input bin sets, making it highly scalable for large and complex metagenomic datasets.

Anvi’o (Eren et al. 2015) is a platform for interactive visualization and manual refinement of metagenomic bins. While it can run automated binning (defaulting to CONCOCT), its main strength lies in allowing users to:

• Inspect contig-level coverage, GC content, and single-copy gene presence
• Visualize connections between contigs in a network view
• Manually merge, split, or reassign contigs to improve bin completeness and reduce contamination
• Annotate bins and link them to taxonomic or functional information

This interactive approach is particularly useful when automated binning produces ambiguous or low-quality bins, enabling high-confidence MAG reconstruction.

Mock binning dataset for this training

Read mapping and binning real metagenomic datasets is a computationally demanding and time-consuming task. To demonstrate the basics of binning in this tutorial, we generated a small mock dataset that is just large enough to produce bins for all binners in this tutorial. The same binners can be applied for any real-life datasets, but as said, plan in some time, up to weeks in some cases.

Agenda

In this tutorial, we will cover:

1. Binning approaches
2. Binning challenges
3. Common binners
4. So many options, what binner to use ?
5. Bin refinement
6. Mock binning dataset for this training
7. Prepare analysis history and data
8. Preparation for binning
9. Binning
   1. Bin contigs using MetaBAT 2
   2. Bin contigs using MaxBin2
   3. Bin contigs using SemiBin
   4. Bin contigs using CONCOCT
   5. Bin contigs using COMEbin
10. Bin refinement
11. Checking the quality of the bins
12. De-replication
13. Conclusions

Prepare analysis history and data

Metagenomic binners typically take two types of data as input:

• A fasta file containing the assembled contigs, which can be generated from raw metagenomic sequencing reads using an assembler such as MEGAHIT, SPAdes, or IDBA-UD.

• A BAM file containing the read coverage information for each contig, which can be generated from the same sequencing reads using mapping software such as Bowtie2 or BWA.

Comment: Can Bins be generated without coverage information?

Not all binners require coverage information. Some, like MetaBAT2, can operate using only genomic composition (e.g., tetranucleotide frequencies) when coverage files are not available. This is especially useful for single-sample datasets or legacy data where coverage cannot easily be calculated.

Other tools that support composition-only binning include:

• MaxBin 2, which can run with composition alone, but performs better with depth.
• SolidBin, which supports single-sample binning based on sequence features.
• VAMB, which primarily uses deep learning on composition, coverage optional.

That said, including coverage information generally increases binning accuracy, especially for:

• Differentiating closely related strains
• Datasets with uneven abundance
• Multi-sample metagenomics workflows (e.g., differential coverage binning)

In summary: yes, it’s possible to bin without coverage, but coverage-aware workflows are recommended when available, as they reduce contamination and improve completeness.

To run binning, we first need to get the data into Galaxy. Any analysis should get its own Galaxy history. So let’s start by creating a new one:

Hands On: Prepare the Galaxy history

1. Create a new history for this analysis

   Tip: Creating a new history

   To create a new history simply click the new-history icon at the top of the history panel:

   

2. Rename the history

   Tip: Renaming a history

   1. Click on galaxy-pencil (Edit) next to the history name (which by default is “Unnamed history”)
   2. Type the new name
   3. Click on Save
   4. To cancel renaming, click the galaxy-undo “Cancel” button

   If you do not have the galaxy-pencil (Edit) next to the history name (which can be the case if you are using an older version of Galaxy) do the following:

   1. Click on Unnamed history (or the current name of the history) (Click to rename history) at the top of your history panel
   2. Type the new name
   3. Press Enter

We need to get the data into our history.

In case of a not very large dataset it’s more convenient to upload data directly from your computer to Galaxy.

Hands On: Upload data into Galaxy

1. Import the contig file from Zenodo or a data library:

   https://zenodo.org/records/17660820/files/MEGAHIT_contigs.fasta
   

   Tip: Importing via links

   • Copy the link location

   • Click galaxy-upload Upload at the top of the activity panel

   • Select galaxy-wf-edit Paste/Fetch Data

   • Paste the link(s) into the text field

   • Press Start

   • Close the window

   Tip: Importing data from a data library

   As an alternative to uploading the data from a URL or your computer, the files may also have been made available from a shared data library:

   1. Go into Libraries (left panel)
   2. Navigate to the correct folder as indicated by your instructor.
      • On most Galaxies tutorial data will be provided in a folder named GTN - Material –> Topic Name -> Tutorial Name.
   3. Select the desired files
   4. Click on Add to History galaxy-dropdown near the top and select as Datasets from the dropdown menu

   5. In the pop-up window, choose

      • “Select history”: the history you want to import the data to (or create a new one)
   6. Click on Import

   Comment

   In case of large dataset, we can use FTP server or the Galaxy Rule-based Uploader.

2. Create a collection named Contigs

   Tip: Creating a dataset collection

   • Click on galaxy-selector Select Items at the top of the history panel
   • Check all the datasets in your history you would like to include

   • Click n of N selected and choose Advanced Build List

     

   • You are in collection building wizard. Choose Flat List and click ‘Next’ button at the right bottom corner.

     

   • Double clcik on the file names to edit. For example, remove file extensions or common prefix/suffixes to cleanup the names.

     

   • Enter a name for your collection
   • Click Build to build your collection
   • Click on the checkmark icon at the top of your history again

3. Import the raw reads in fastq format from Zenodo or a data library:

   https://zenodo.org/records/17660820/files/reads_forward.fastqsanger.gz
   https://zenodo.org/records/17660820/files/reads_reverse.fastqsanger.gz
   

4. Create a paired collection named Reads

   Tip: Creating a paired collection

   • Click on galaxy-selector Select Items at the top of the history panel
   • Check all the datasets in your history you would like to include

   • Click n of N selected and choose Advanced Build List

     

   • You are in collection building wizard. Choose Flat List and click ‘Next’ button at the right bottom corner.

     

   • Check and configure auto-pairing. Commonly matepairs have suffix _1 and _2 or _R1 and _R2. Click on ‘Next’ at the bottom.

     

   • Edit the List Identifier as required.
   • Enter a name for your collection
   • Click Build to build your collection
   • Click on the checkmark icon at the top of your history again

Comment: Why do we use collections here?

In this tutorial, collections are not strictly necessary because we are working with only one contig file and its paired-end reads. However, in real metagenomic studies, it is common to process many samples—sometimes hundreds or even thousands—and in those cases, collections become essential for managing data efficiently.

It is generally good practice to first test a workflow on a small subset of the data (for example, a collection containing only a single sample) to ensure that the tools run correctly and the parameters are appropriate before launching thousands of jobs on Galaxy.

Preparation for binning

As explained before, we need coverage information in a BAM file as a requirement for all binners. Some binners need a specific format for the coverage information, but this will be covered in the version specific to the desired binner. For now, we will map the quality-controlled reads to the contigs to get a BAM file with the coverage information. This BAM file also needs to be sorted for the downstream binners.

Comment

Make sure the reads are quality-controlled, e.g., by following the Quality Control tutorial

Hands On: Map reads to contigs

1. Bowtie2 ( Galaxy version 2.5.4+galaxy0) with the following parameters:
   • “Is this single or paired library”: Paired-end
     • param-collection “FASTQ Paired Dataset”: Reads (Input dataset collection)
     • “Do you want to set paired-end options?”: No
   • “Will you select a reference genome from your history or use a built-in index?”: Use a genome from the history and build index
     • param-collection “Select reference genome”: Contigs (Input dataset collection)
   • “Set read groups information?”: Do not set
   • “Select analysis mode”: 1: Default setting only
   • “Do you want to tweak SAM/BAM Options?”: No
   • “Save the bowtie2 mapping statistics to the history”: Yes

Let’s now sort the BAM files.

Hands On: Sort BAM files

1. Samtools sort ( Galaxy version 2.0.7) with the following parameters:
   • param-file “BAM File”: output of Bowtie2 tool
   • “Primary sort key”: coordinate

The sorted BAM file can be used as input for any of the binning tools.

Binning

As explained before, there are many challenges to metagenomics binning. The most common of them are listed below:

• High complexity.
• Fragmented sequences.
• Uneven coverage.
• Incomplete or partial genomes.
• Horizontal gene transfer.
• Chimeric sequences.
• Strain variation.

Open image in new tab

In this tutorial, we offer several versions, each highlighting a different binners:

Hands-on: Choose Your Own Tutorial

This is a "Choose Your Own Tutorial" (CYOT) section (also known as "Choose Your Own Analysis" (CYOA)), where you can select between multiple paths. Click one of the buttons below to select how you want to follow the tutorial

MetaBAT2 MaxBin2 SemiBin CONCOCT COMEBin

Bin refinement

Now that we have produced bins with our favorite binner, we can refine the recovered bins. Therefore, we need to

1. Convert the bins into a contig-to-bin mapping table,
2. Combine the tables from each binner into one collection, and

3. Use Binette (Mainguy and Hoede 2024) to create consensus bins.

   An alternative tool would be DAS Tool ( Galaxy version 1.1.7+galaxy1) (Sieber et al. 2018) which is also available in Galaxy.

For the refinement, we will use the bins created by all the binners used before. If you did not run them all by yourself, we have provided the results here as well.

Hands On: Get result bins

1. Import the bin files from Zenodo or a data library:

   https://zenodo.org/records/17660820/files/semibin_0.fasta
   https://zenodo.org/records/17660820/files/maxbin_0.fasta
   https://zenodo.org/records/17660820/files/maxbin_1.fasta
   https://zenodo.org/records/17660820/files/metabat_0.fasta
   https://zenodo.org/records/17660820/files/concoct_1.fasta
   https://zenodo.org/records/17660820/files/concoct_2.fasta
   https://zenodo.org/records/17660820/files/concoct_3.fasta
   https://zenodo.org/records/17660820/files/concoct_4.fasta
   https://zenodo.org/records/17660820/files/concoct_5.fasta
   https://zenodo.org/records/17660820/files/concoct_6.fasta
   https://zenodo.org/records/17660820/files/concoct_7.fasta
   https://zenodo.org/records/17660820/files/concoct_8.fasta
   https://zenodo.org/records/17660820/files/concoct_9.fasta
   

2. Create a collection for each binner by selecting only the bins created by this binner (e.g., maxbin_0.fasta and maxbin_1.fasta together) and creating a collection:

   Tip: Creating a dataset collection

   • Click on galaxy-selector Select Items at the top of the history panel
   • Check all the datasets in your history you would like to include

   • Click n of N selected and choose Advanced Build List

     

   • You are in collection building wizard. Choose Flat List and click ‘Next’ button at the right bottom corner.

     

   • Double clcik on the file names to edit. For example, remove file extensions or common prefix/suffixes to cleanup the names.

     

   • Enter a name for your collection
   • Click Build to build your collection
   • Click on the checkmark icon at the top of your history again

Once each bin set is converted into one collection, it can be converted into a contig-to-bin mapping table. We need to perform this step for every bin set.

Hands On: Convert the bins into a contig-to-bin mapping table

1. Converts genome bins in fasta format ( Galaxy version 1.1.7+galaxy1) with the following parameters:
   • param-file “Bin sequences”: bins (output of any of the binners tool)

We can now build a list of the tables

Hands On: Build a list of the binning tables

1. Build list with the following parameters:
   • In “Dataset”:
     • param-repeat “Insert Dataset”
       • param-file “Input Dataset”: contigs2bin (output of Converts genome bins in fasta format of SemiBin tool)
       • “Label to use”: Index
     • param-repeat “Insert Dataset”
       • param-file “Input Dataset”: contigs2bin (output of Converts genome bins in fasta format of MetaBAT2 tool)
       • “Label to use”: Index
     • param-repeat “Insert Dataset”
       • param-file “Input Dataset”: contigs2bin (output of Converts genome bins in fasta format of MaxBin2 tool)
       • “Label to use”: Index
     • param-repeat “Insert Dataset”
       • param-file “Input Dataset”: contigs2bin (output of Converts genome bins in fasta format of CONCOCT tool)
       • “Label to use”: Index

We can now refine the bins with Binette.

Hands On: Refine with Binette

1. Binette ( Galaxy version 1.2.0+galaxy0) with the following parameters:
   • param-file “Input contig table”: output (output of Build list tool)
   • param-collection “Input contig file”: output (Input dataset collection)
   • “Select if database should be used either via file or cached database”: cached database

Question: Bin refinement

1. How many bins are left after refinement ?

Solution

1. Two bins are left. Most contigs from different bins where combined into one bin. There is still one single contig bin left.

Checking the quality of the bins

Once binning is done, it is important to check its quality.

Binning results can be evaluated with CheckM (Parks et al. 2015). CheckM is a software tool used in metagenomics binning to assess the completeness and contamination of genome bins.

CheckM compares the genome bins to a set of universal single-copy marker genes that are present in nearly all bacterial and archaeal genomes. By identifying the presence or absence of these marker genes in the bins, CheckM can estimate the completeness of each genome bin (i.e., the percentage of the total set of universal single-copy marker genes that are present in the bin) and the degree of contamination (i.e., the percentage of marker genes that are found in more than one bin).

This information can be used to evaluate the quality of genome bins and to select high-quality bins for further analysis, such as genome annotation and comparative genomics. CheckM is widely used in metagenomics research and has been shown to be an effective tool for assessing the quality of genome bins. Some of the key functionalities of CheckM are:

• Estimation of genome completeness: CheckM uses a set of universal single-copy marker genes to estimate the completeness of genome bins. The completeness score indicates the proportion of these marker genes that are present in the bin, providing an estimate of how much of the genome has been recovered.

• Estimation of genome contamination: CheckM uses the same set of marker genes to estimate the degree of contamination in genome bins. The contamination score indicates the proportion of marker genes that are present in multiple bins, suggesting that the genome bin may contain DNA from more than one organism.

• Identification of potential misassemblies: CheckM can identify potential misassemblies in genome bins based on the distribution of marker genes across the genome.

• Visualization of results: CheckM can generate various plots and tables to visualize the completeness, contamination, and other quality metrics for genome bins, making it easier to interpret the results.

• Taxonomic classification: CheckM can also be used to classify genome bins taxonomically based on the presence of specific marker genes associated with different taxonomic groups.

Based on the previous analysis we will use CheckM lineage_wf: Assessing the completeness and contamination of genome bins using lineage-specific marker sets

CheckM lineage_wf is a specific workflow within the CheckM software tool that is used for taxonomic classification of genome bins based on their marker gene content. This workflow uses a reference database of marker genes and taxonomic information to classify the genome bins at different taxonomic levels, from domain to species.

Now we can investigate the completeness and contamination of any of your previously generated genome bins, as well as the refined set.

Hands On: Assessing the completeness and contamination of genome bins using lineage-specific marker sets with `CheckM lineage_wf`

1. CheckM lineage_wf ( Galaxy version 1.2.0+galaxy0) with parameters:
   • “Bins”: Folder containing the produced bins

The output of “CheckM lineage_wf” includes several files and tables that provide information about the taxonomic classification and quality assessment of genome bins. Here are some of the key outputs:

• CheckM Lineage Workflow Output Report (Bin statistics): This report provides a summary of the quality assessment performed by CheckM. It includes statistics such as the number of genomes analyzed, their completeness, contamination, and other quality metrics.

• Lineage-specific Quality Assessment: CheckM generates lineage-specific quality assessment files for each analyzed genome. These files contain detailed information about the completeness and contamination of the genome based on its taxonomic lineage.

• Marker Set Analysis: CheckM uses a set of marker genes to estimate genome completeness and contamination. The tool produces marker-specific analysis files that provide details on the presence, absence, and copy number of each marker gene in the analyzed genomes.

• Visualizations: CheckM generates various visualizations to aid in the interpretation of the results. These include plots such as the lineage-specific completeness and contamination plots, scatter plots, and other visual representations of the data.

• Tables and Data Files: CheckM generates tabular data files that contain detailed information about the analyzed genomes, including their names, taxonomic assignments, completeness scores, contamination scores, and other relevant metrics. These files are useful for further downstream analysis or data manipulation.

It should be noted that “CheckM lineage_wf” offers a range of optional outputs that can be generated to provide additional information to the user.

To keep it simple, we will check the bin statistics to investigate the performance of our previous binning efforts.

Question: Binning performance

1. Which binner created the best bins?
2. How well did the bin refinement work?

Solution

1. Let’s combine the stat:

   Binner | Bin ID | Marker lineage | # genomes | # markers | # marker sets | 0 | 1 | 2 | 3 | 4 | 5+ | Completeness | Contamination | Strain heterogeneity — | — | — | — | — | — | — | — | —| — | — | — | — | — | — MetaBAT2 | 1 | k__Bacteria (UID203) | 5449 | 103 | 58 | 89 | 14 | 0 | 0 | 0 | 0 | 15.67 | 0.00 0.00 SemiBin | SemiBin_0 | k__Bacteria (UID203) | 5449 | 103 | 58 | 89 | 14 | 0 | 0 | 0 | 0 | 15.67 | 0.00 | 0.00 MaxBin | 001 | k__Bacteria (UID203) | 5449 | 103 | 58 | 92 | 11 | 0 | 0 | 0 | 0 | 10.50 | 0.00 | 0.00 MaxBin | 002 | k__Bacteria (UID203) | 5449 | 103 | 58 | 99 | 4 | 0 | 0 | 0 | 0 | 6.90| 0.00 | 0.00 CONCOCT | 0 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 1 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 2 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 3 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 4 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 5 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 6 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 7 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 8 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00 CONCOCT | 9 | k__Bacteria (UID203) |5449 | 103 | 58 | 89 | 14 | 0 | 0 | 0 | 0 | 15.67 | 0.00 | 0.00 MetaBAT2 and SemiBin generated the best‐performing bins in this dataset, each recovering one bin with ~15.7% completeness and no detectable contamination.

   MaxBin produced bins of lower completeness (10.5% and 6.9%), but still without contamination.

   CONCOCT largely failed to recover meaningful bins in this dataset: most bins show 0% completeness. This often occurs with CONCOCT on small assemblies or uneven coverage, because its Gaussian clustering model struggles when the number of contigs is low or the variation in coverage is insufficient. Binette

| Bin ID | Marker lineage | # genomes | # markers | # marker sets | 0 | 1 | 2 | 3 | 4 | 5+ | Completeness | Contamination | Strain heterogeneity | |—————|————————|———–|————|—————-|—-|—-|—-|—-|—-|—-|————–|—————|———————–| | binette_bin1 | k__Bacteria (UID203) | 5449 | 103 | 58 | 89 | 14 | 0 | 0 | 0 | 0 | 15.67 | 0.00 | 0.00 |

1. Let’s look at Binette stats: Bin ID | Marker lineage | # genomes | # markers | # marker sets | 0 | 1 | 2 | 3 | 4 | 5+ | Completeness | Contamination | Strain heterogeneity — | — | — | — | — | — | — | —| — | — | — | — | — | — binette_bin1 | k__Bacteria (UID203) | 5449 | 103 | 58 | 89 | 14 | 0 | 0 | 0 | 0 | 15.67 | 0.00 | 0.00 binette_bin2 | root (UID1) | 5656 | 56 | 24 | 56 | 0 | 0 | 0 | 0 | 0 | 0.00 | 0.00 | 0.00

Binette produced one bin essentially identical in quality to the MetaBAT2/SemiBin bins (~15.7% completeness), plus a root-level bin with no completeness. Binette could recover the most completely successful bin. The additional loq quality bin, would have normally been filtered by Binette if the Set minimus completeness parameter were set to a reasonable value, often > 75 % for true biological data.

Comment: CheckM2 vs CheckM

CheckM2 (Chklovski et al. 2023) is the successor of CheckM, but CheckM is still widely used, since its marker-based logic can be more interpretable in a biological sense. E.g., to date (2025-11-21), NCBI still allows submitting MAGs to GenBank if either checkM or checkM2 has a completeness of > 90% (see the NCBI WGS/MAG submission guidelines).

Key differences compared to CheckM1:

• CheckM1 relies primarily on lineage-specific single-copy marker genes to estimate completeness and contamination of microbial genomes.
• CheckM2 uses a machine-learning (gradient boost / ML) approach trained on simulated and experimental genomes, and does not strictly require a well-represented lineage in its marker database.
• CheckM2 is reported to be more accurate and faster for both bacterial and archaeal lineages, especially when dealing with novel or very reduced-genome lineages (e.g., candidate phyla, CPR/DPANN) where classical marker-gene methods may struggle.
• The database of CheckM2 can be updated more rapidly with new high-quality reference genomes, which supports scalability and improved performance over time.

If you’re working with MAGs from underrepresented taxa (novel lineages) or very small genomes (streamlined bacteria/archaea), CheckM2 tends to give more reliable estimates of completeness/contamination. For more “standard” microbial genomes from well-studied taxa, CheckM1 may still work well, but you may benefit from the improved performance with CheckM2.

De-replication

De-replication is the process of identifying sets of genomes that are the “same” in a list of genomes, and removing all but the “best” genome from each redundant set. How similar genomes need to be to be considered “same”, how to determine which genome is “best”, and other important decisions are discussed in Important Concepts.

A common use for genome de-replication is the case of individual assembly of metagenomic data. If metagenomic samples are collected in a series, a common way to assemble the short reads is with a “co-assembly”. That is, combining the reads from all samples and assembling them. The problem with this is that assembling similar strains can severely fragment assemblies, precluding the recovery of a good genome bin. An alternative option is to assemble each sample separately, and then “de-replicate” the bins from each assembly to make a final genome set.

Open image in new tab

Several tools have been designed for the process of de-replication. dRep (Olm et al. 2017) is a software tool designed for the dereplication of genomes in metagenomic datasets. The goal is to retain a representative set of genomes to improve downstream analyses, such as taxonomic profiling and functional annotation.

An typical workflow of how dRep works for dereplication in metagenomics includes:

• Genome Comparison: dRep uses a pairwise genome comparison approach to assess the similarity between genomes in a given metagenomic dataset.

• Clustering: Based on the genome similarities, dRep performs clustering to group similar genomes into “genome clusters.” Each cluster represents a group of closely related genomes.

• Genome Quality Assessment: dRep evaluates the quality of each genome within a cluster. It considers factors such as completeness, contamination, and strain heterogeneity.

• Genome Selection: Within each genome cluster, dRep selects a representative genome based on user-defined criteria. This representative genome is considered the “dereplicated” version of the cluster.

• Dereplication Output: The output of dRep includes information about the dereplicated genomes, including their identity, completeness, and contamination. The user can choose a threshold for genome similarity to control the level of dereplication.

Hands On: General list of actions for de-replication

1. Create new history
2. Assemble each sample separately using your favorite assembler
3. Perform a co-assembly to catch low-abundance microbes
4. Bin each assembly separately using your favorite binner
5. Bin co-assembly using your favorite binner
6. Pull the bins from all assemblies together
7. Run dRep dereplication ( Galaxy version 3.6.2+galaxy1) on them
8. Perform downstream analysis on the de-replicated genome list

Conclusions

In summary, this tutorial shows a step-by-step guide on how to bin metagenomic contigs using various binners, including bin refinement.

It is critical to select the appropriate binning tool for a specific metagenomics study, as different binning methods may have different strengths and limitations depending on the type of metagenomic data being analyzed. By comparing the outcomes of several binning techniques, researchers can increase the precision and accuracy of genome binning.

There are various binning methods available for metagenomic data, including reference-based, clustering-based, hybrid approaches, and machine learning. Each method has its advantages and disadvantages, and the selection of the appropriate method depends on the research question and the characteristics of the data.

Comparing the outcomes of multiple binning methods can help to identify the most accurate and reliable method for a specific study. This can be done by evaluating the quality of the resulting bins in terms of completeness, contamination, and strain heterogeneity, as well as by comparing the composition and functional profiles of the identified genomes.

Overall, by carefully selecting and comparing binning methods, researchers can improve the quality and reliability of genome bins, which can ultimately lead to a better understanding of the functional and ecological roles of microbial communities in various environments.

You've Finished the Tutorial

Please also consider filling out the Feedback Form as well!

Key points

• Metagenomics binning is a computational approach to grouping together DNA sequences from a mixed microbial sample into metagenome-assembled genomes (MAGs)

• The metagenomics binning workflow involves several steps, including preprocessing of raw sequencing data, assembly of sequencing reads into contigs, binning of contigs into MAGs, quality assessment of MAGs, and annotation of functional genes and metabolic pathways in MAGs

• The quality and completeness of MAGs can be evaluated using standard metrics, such as completeness, contamination, and genome size

• Metagenomics binning can be used to gain insights into the composition, diversity, and functional potential of microbial communities, and can be applied to a range of research areas, such as human health, environmental microbiology, and biotechnology

Frequently Asked Questions

• Topic FAQs
• Galaxy FAQs
• GTN Matrix Chat
• Galaxy Help Forum

Useful literature

Further information, including links to documentation and original publications, regarding the tools, analysis techniques and the interpretation of results described in this tutorial can be found here.

References

1. Alneberg, J., B. S. Bjarnason, I. de Bruijn, M. Schirmer, J. Quick et al., 2014 Binning metagenomic contigs by coverage and composition. Nature Methods 11: 1144–1146. 10.1038/nmeth.3103
2. Imelfort, M., D. Parks, B. J. Woodcroft, P. Dennis, P. Hugenholtz et al., 2014 GroopM: an automated tool for the recovery of population genomes from related metagenomes. PeerJ 2: e603. 10.7717/peerj.603
3. Eren, A. M., Özcan C. Esen, C. Quince, J. H. Vineis, H. G. Morrison et al., 2015 Anvi’o: an advanced analysis and visualization platform for ‘omics data. PeerJ 3: e1319. 10.7717/peerj.1319
4. Parks, D. H., M. Imelfort, C. T. Skennerton, P. Hugenholtz, and G. W. Tyson, 2015 CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Research 25: 1043–1055. 10.1101/gr.186072.114
5. Wu, Y.-W., B. A. Simmons, and S. W. Singer, 2015 MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics 32: 605–607. 10.1093/bioinformatics/btv638
6. Lin, H.-H., and Y.-C. Liao, 2016 Accurate binning of metagenomic contigs via automated clustering sequences using information of genomic signatures and marker genes. Scientific Reports 6: 10.1038/srep24175
7. Olm, M. R., C. T. Brown, B. Brooks, and J. F. Banfield, 2017 dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. The ISME journal 11: 2864–2868. 10.1038/ismej.2017.126
8. Sieber, C. M. K., A. J. Probst, A. Sharrar, B. C. Thomas, M. Hess et al., 2018 Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy. Nature Microbiology 3: 836–843. 10.1038/s41564-018-0171-1
9. Uritskiy, G. V., J. DiRuggiero, and J. Taylor, 2018 MetaWRAP—a flexible pipeline for genome-resolved metagenomic data analysis. Microbiome 6: 10.1186/s40168-018-0541-1
10. Kang, D. D., F. Li, E. Kirton, A. Thomas, R. Egan et al., 2019 MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ 7: e7359. 10.7717/peerj.7359
11. Nissen, J. N., J. Johansen, R. L. Allesøe, C. K. Sønderby, J. J. A. Armenteros et al., 2021 Improved metagenome binning and assembly using deep variational autoencoders. Nature biotechnology 39: 555–560. 10.1038/s41587-020-00777-4
12. Meyer, F., A. Fritz, Z. L. Deng, D. Koslicki, T. R. Lesker et al., 2022 Critical Assessment of Metagenome Interpretation: the second round of challenges. Nature Methods 19: 429–440. 10.1038/s41592-022-01431-4
13. Pan, S., C. Zhu, X.-M. Zhao, and L. P. Coelho, 2022 A deep siamese neural network improves metagenome-assembled genomes in microbiome datasets across different environments. Nature Communications 13: 10.1038/s41467-022-29843-y
14. Chklovski, A., D. H. Parks, B. J. Woodcroft, and G. W. Tyson, 2023 CheckM2: a rapid, scalable and accurate tool for assessing microbial genome quality using machine learning. Nature Methods 20: 1203–1212. 10.1038/s41592-023-01940-w
15. Mainguy, J., and C. Hoede, 2024 Binette: a fast and accurate bin refinement tool to construct high‐quality Metagenome Assembled Genomes. Journal of Open Source Software 9: 6782. 10.21105/joss.06782
16. Wang, Z., R. You, H. Han, W. Liu, F. Sun et al., 2024 Effective binning of metagenomic contigs using contrastive multi‑view representation learning. Nature Communications 15: 10.1038/s41467-023-44290-z
17. Author, A., B. Author, and C. Author, 2025 Comprehensive benchmarking of metagenomic binners across diverse environments. Nature Communications 16: 57957. 10.1038/s41467-025-57957-6

Feedback

Did you use this material as an instructor? Feel free to give us feedback on how it went.
Did you use this material as a learner or student? Click the form below to leave feedback.

Citing this Tutorial

1. Paul Zierep, Nikos Pechlivanis, Fotis E. Psomopoulos, Vini Salazar, Binning of metagenomic sequencing data (Galaxy Training Materials). https://training.galaxyproject.org/training-material/topics/microbiome/tutorials/metagenomics-binning/tutorial.html Online; accessed TODAY
2. Hiltemann, Saskia, Rasche, Helena et al., 2023 Galaxy Training: A Powerful Framework for Teaching! PLOS Computational Biology 10.1371/journal.pcbi.1010752
3. Batut et al., 2018 Community-Driven Data Analysis Training for Biology Cell Systems 10.1016/j.cels.2018.05.012

BibTeX

@misc{microbiome-metagenomics-binning,
author = "Paul Zierep and Nikos Pechlivanis and Fotis E. Psomopoulos and Vini Salazar",
	title = "Binning of metagenomic sequencing data (Galaxy Training Materials)",
	year = "",
	month = "",
	day = "",
	url = "\url{https://training.galaxyproject.org/training-material/topics/microbiome/tutorials/metagenomics-binning/tutorial.html}",
	note = "[Online; accessed TODAY]"
}
@article{Hiltemann_2023,
	doi = {10.1371/journal.pcbi.1010752},
	url = {https://doi.org/10.1371%2Fjournal.pcbi.1010752},
	year = 2023,
	month = {jan},
	publisher = {Public Library of Science ({PLoS})},
	volume = {19},
	number = {1},
	pages = {e1010752},
	author = {Saskia Hiltemann and Helena Rasche and Simon Gladman and Hans-Rudolf Hotz and Delphine Larivi{\`{e}}re and Daniel Blankenberg and Pratik D. Jagtap and Thomas Wollmann and Anthony Bretaudeau and Nadia Gou{\'{e}} and Timothy J. Griffin and Coline Royaux and Yvan Le Bras and Subina Mehta and Anna Syme and Frederik Coppens and Bert Droesbeke and Nicola Soranzo and Wendi Bacon and Fotis Psomopoulos and Crist{\'{o}}bal Gallardo-Alba and John Davis and Melanie Christine Föll and Matthias Fahrner and Maria A. Doyle and Beatriz Serrano-Solano and Anne Claire Fouilloux and Peter van Heusden and Wolfgang Maier and Dave Clements and Florian Heyl and Björn Grüning and B{\'{e}}r{\'{e}}nice Batut and},
	editor = {Francis Ouellette},
	title = {Galaxy Training: A powerful framework for teaching!},
	journal = {PLoS Comput Biol}
}

                   

Galaxy Administrators: Install the missing tools

You can use Ephemeris's shed-tools install command to install the tools used in this tutorial.

shed-tools install [-g GALAXY] [-a API_KEY] -t <(curl https://training.galaxyproject.org/training-material/api/topics/microbiome/tutorials/metagenomics-binning/tutorial.json | jq .admin_install_yaml -r)

Alternatively you can copy and paste the following YAML

---
install_tool_dependencies: true
install_repository_dependencies: true
install_resolver_dependencies: true
tools:
- name: checkm_lineage_wf
  owner: iuc
  revisions: f0107b9f2dc3
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: concoct
  owner: iuc
  revisions: eae7ee167917
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: concoct_coverage_table
  owner: iuc
  revisions: fd31cd168efc
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: concoct_cut_up_fasta
  owner: iuc
  revisions: 4d8bc5dd9e95
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: concoct_extract_fasta_bins
  owner: iuc
  revisions: 8b1b09fcd8b7
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: concoct_merge_cut_up_clustering
  owner: iuc
  revisions: 20ccec4a2c38
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: das_tool
  owner: iuc
  revisions: b048a987dd7d
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: drep_dereplicate
  owner: iuc
  revisions: f54e7b3da33c
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: fasta_to_contig2bin
  owner: iuc
  revisions: fb2bed0eb02f
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: metabat2
  owner: iuc
  revisions: f375b4f6ef57
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: metabat2_jgi_summarize_bam_contig_depths
  owner: iuc
  revisions: 00e3b4ef7e0c
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/
- name: maxbin2
  owner: mbernt
  revisions: 9da9706f7ca9
  tool_panel_section_label: Metagenomic Analysis
  tool_shed_url: https://toolshed.g2.bx.psu.edu/

Feedback

5 stars	3
4 stars	2
3 stars	1
1 stars	2

May 2025

• 1 stars: Liked: The introduction was good. Disliked: Following the instructions would not produce expected results. MetaBat2 will not produce bin reads if cluster membership is set to 'Yes'. The instructions for running CheckM is a repeat of MetaBAT2 instructions.
• 4 stars: Liked: quick tutorial, learnt something new :) Disliked: 1) Import generated assembly files: these files could not be downloaded (typo ??) Had to get the files in an alternate way. 2) The settings of "Checkm lineage_wf" as described above do not fit the tool's settings.

April 2025

• 4 stars: Liked: I like that it provides an overview of the project with enough detail to get started with an analysis. The links to the more detailed information makes it a great step into understanding the process and not just blindly following the "map".

October 2024

• 5 stars: Liked: hte flow of the tutorial Disliked: the tutorial is comprehensive more notes on intepreting results