Skip to content

Latest commit

 

History

History
67 lines (57 loc) · 4.69 KB

File metadata and controls

67 lines (57 loc) · 4.69 KB

Metagenomic binning

Binning is essentially clustering for assembled contigs. Create draft metagenome-assembled genomes and evaluate their completeness, contamination and other metrics with these helpful tools!

This pipeline uses DASTool, which aggregates results from several binning methods to produce a higher-quality bin set. The DASTool pipeline here uses three binners under the hood: Metabat2, Maxbin and CONCOCT. While we chose these binners and DASTool, there are several other binners and aggregation tools out there in the literature.

Kraken2 is currently used to make a taxonomic assignment of a bin. You need a functional Kraken2 database for this pipeline to work.

Running DASTool

This pipeline will work for a set of many samples at once. Copy the binning/config_binning_manysamp.yaml to your working directory and change the parameters.

Specify the following fields in the configfile (defaults are provided) - outdir_base (a folder for each sample will be created within here) - sample_file (see below) - read_length - kraken2db - custom_taxonomy - long_read

Your sample file must be a tab delimited file with three columns and one row for each sample. Forward and reverse reads should be in the same column separated by commas.

SAMPLE_NAME    ASSEMBLY   READS1,READS2

Then, you can launch the workflow to submit jobs to the SCG SLURM cluster. Note the use of --use-conda on the end to get certain rules to work correctly.

snakemake --configfile config_binning_manysamp.yaml --snakefile path/to/bin_das_tool_manysamp.snakefile \
--profile scg --jobs 999 --use-singularity --singularity-args '--bind /oak/,/labs/,/home' --use-conda

Comparing bins against references

THIS PIPELINE HAS BEEN DEPRECATED WITHIN THE BHATT LAB, I DELETED THE DATABASES To compare your freshly minted bins against reference genomes from Genbank and MAG databases, run the compare_bins_references.snakefile pipeline. For each sample, this will compare the bin against the genbank and MAG databases with MASH. The top 50 matches will be passed into fastani, and the top 10 from there will get more careful nucmer alignments and dotplots. This pipeline relies on the directory structure output from the binning pipeline. The file config_compare_bins_references.yaml takes two inputs:

  • base output directory from binnnig. expects one folder for each sample within this directory. This should be where you ran DASTool or Metabat
  • sample file: one column, newline delimited for each sample this is just the sample name. There should be a folder in outdir_base for each sample name specified here.

Running on Nanopore assemblies

Change the long_read specification in the configfile and put your nanopore reads in the third column (just once). If you also have high-quality short read data, you could run this pipeline in the standard mode by providing your nanopore assembly and the short read pairs. This might be more accurate for depth/coverage calculations, but I'm not sure.

Still in development, please file issues on GitHub if you encounter any.

Result metadata

The best description of what each bin will be in the binning_table_all_simple.tsv file. A description of the meanings of each column is below.

Sample                 sample name
Bin                    bin name, corresponds to fasta file
bin.quality.numeric    Completeness – (5*contamination)
bin.quality.call       0-4 quality call, based on Nayfach and Bowers standards
Completeness           completeness from checkM
Contamination          contamination from checkM
Strain.heterogeneity   Strain heterogeneity from checkM
lca_species            kraken2-based bin classifiction. At least 2/3 total bin length was classified to this taxa
lca_level              taxonomic level of the above taxa
lca_fraction           what fraction of total length was classified as the above taxa
best_species           kraken2-based bin classifiction. Single species with the highest abundance
best_level             taxonomic level of the above taxa
best_fraction          what fraction of total length was classified as the above taxa
N50                    Bin N50 (contiguity)
Size.Mb                size of bin in Mb
Coverage               Coverage of reads mapped back to bin
# contigs (>= 0 bp)    Total number of contigs in bin
Largest contig         largest contig in bin
Genes                  number of genes predicted by prodigal
tRNA                   number of tRNA genes predicted by prokka
rna.16S                number of 16S genes predicted by prokka
rna.23S                number of 23S genes predicted by prokka
rna.5S                 number of 5S genes predicted by prokka