Metagenome-Assembled Genomes (MAGs) Generation

This workflow generates Metagenome-Assembled Genomes (MAGs) from paired short reads.
Dereplicated MAGs for the complete input sample set are reported.

Workflow Logic

The workflow supports assembly using metaSPADES and MEGAHIT.
For binning, it utilizes four different tools: MetaBAT2, MaxBin2, SemiBin, and CONCOCT. The resulting bins are then refined using Binette, the successor of metaWRAP.

MAGs Annotation and Quality Control

After binning, the resulting MAGs are dereplicated across all input samples based on CheckM2 quality metrics using dRep. The following processing steps are then performed:

• Annotation with Bakta
• Taxonomic Assignment using GTDB-Tk
• Quality Control via QUAST and CheckM/CheckM2
• Abundance Estimation per sample with CoverM

All results are consolidated into a single MultiQC report for easy analysis.

Input Requirements

Input reads must be quality-filtered, with host reads removed. See other MAGs workflows in IWC for quality filtering and host removal.

• Trimmed reads: Quality-trimmed reads from individual samples.
• Trimmed reads from grouped samples: These reads need to be grouped based on the desired MAGs generation approach. The tool fastq_groupmerge can be used to perform the grouping.
  • Individual MAGs Generation: Use the same input as Trimmed Reads to generate MAGs per sample.
  • Pooled MAGs Generation (Co-assembly/Binning): Merge all reads input one file for a fully pooled MAGs approach.
  • Grouped MAGs Generation (Co-assembly/Binning): Merge samples based on predefined groups.
  • Hybrid MAGs Generation: Combine individual and grouped reads for a mixed approach.

Note: Merging reads can result in large input files, significantly increasing computational demands—especially during assembly and binning, which may require substantial RAM. Our tests with synthetic samples up to 50 GB showed feasible performance. For larger datasets, we recommend limiting the approach to individual or pooled MAGs generation.

Optional modifications

Some modifications that can be done via the workflow editor for specific use cases:

• Compare binners / bin refinement: add checkM/M2 to each binner / refiner and compare completeness, contamination, n. of bins.
• Optionally use DAS tool instead of Binette (our benchmark showed that Binette produce better MAGs on average).
• Add annotation for Bakta (currently Bakta only reports rRNAs to save resources), other annotations can be selected in the Bakta options.
• Downstream: compare differential abundance of MAGs in your samples based on metadata using maaslin2.
• Downstream: compare your MAGs to a catalogue of MAGs for similar samples. Download desired MAGs e.g. from MGnify genomes you can query your MAGs via the MGnify API; download the Genomes and compare in Galaxy e.g. using dREP.