About:

I built this snakemake pipeline to showcase how FASTQ files can be taken all the way into a set of good-quality dereplicated MAGs.

The general steps are:

• Quality checked the reads using fastQC
• Assembled the fastq reads using SPADES
• Bin the MAGs using metawrap
• Refining MAGs using dasTool
• Deplicate the MAGs (if relevant) using dRep.
• Determine MAG quality using checkM
• Select only MIMAG quality-standard MAGs for further analyses (e.g. >50% complete, <10% contamination).
• Assign taxonomy of this MAG set using gtdbtk.

Example result folder:

https://github.com/patriciatran/MAG_pipeline/blob/main/example_results_folder.txt

Relevant folders for output:

• results/{sample}/final_bin_set/.fasta* : all the final bins in FASTA format
• results/{sample}/taxonomy_final_bin_set.tsv : final GTDBTK taxonomic assignment for the final bin set of MAGs

Status:

April 17, 2023

• Pipeline works without errors!
• Next step: improving documentation and distribute as a package.
• Add ways to report final information : e.g. run time of the pipeline, how many MAGs in the final bin set for each sample.
• Add ways to report final information: e.g. bar plot of taxonomies across samples

Thanks to:

This pipeline exists because of the folks making these programs available, please cite their work:

• SPADES: https://github.com/ablab/spades

• Metawrap: https://github.com/bxlab/metaWRAP

• Metabat1 and Metabat2: https://bitbucket.org/berkeleylab/metabat

• Maxbin2: https://sourceforge.net/projects/maxbin/

• DasTool: https://github.com/cmks/DAS_Tool

• dRep: https://github.com/MrOlm/drep

• CheckM: https://github.com/Ecogenomics/CheckM

• GTDBTK: https://github.com/Ecogenomics/GTDBTk

• Snakemake: https://snakemake.readthedocs.io/en/stable/

• conda/anaconda: https://docs.anaconda.com/anaconda/user-guide/faq/

• mamba: https://github.com/mamba-org/mamba

Further Reading:

MIMAG Standards: https://www.nature.com/articles/nbt.3893

Data used for pipeline testing:

Tisza MJ et al., "A catalog of tens of thousands of viruses from human metagenomes reveals hidden associations with chronic diseases.", Proc Natl Acad Sci U S A, 2021 Jun 8;118(23)