Shell Scripts for use with QIIME

These are just very simple shell scripts that wrap the QIIME commands developed by Rob Knights's group. They primarily save time on typing but are particularly helpful in the case of having many datasets, where you can utilize looping, for example:

cd folder/with/files
for file in *; do bash shellscript.sh $file; done

Requirements

• QIIME2-2019.10 and up

Usage

Run as in any shell script, i.e.

bash /relative/filepath/to/shellscript.sh

the first run is an echo command that will prompt for what inputs to utilize

Details about each script

0. To start, import your features x samples .tsv file. and convert to .qza. Note, this .tsv should not have a "taxonomy" column at the end.

bash ../../../fast-16s-analysis/shell_scripts/1-qiime-tools-import.sh Trios_ASV_table_Silva_v138_1.tsv 

1. To extract the counts for samples that belong to specific groups, use filter-ASV-by-metadata.sh $ASV.qza $Metadata $Metadata-column $Metadata-column-value. Here, I'm extracting the counts for the samples which belong only to my mice which fulfill the criteria of "JAX" for "Background".

bash ../../../../fast-16s-analysis/shell_scripts/filter-ASV-by-metadata.sh Baseline_ASV_table_Silva_v138_1.qza Baseline_Metadata.tsv Background JAX

2. To make a taxa summary plot (stacked column chart showing taxonomy)

• Make a taxonomy.tsv file, which should contain the header Feature ID, Taxon
• Import the taxonomy.tsv file with import-taxonomy.sh.

bash ../../../fast-16s-analysis/shell_scripts/import-taxonomy.sh final_taxonomy.tsv 
Imported final_taxonomy.tsv as TSVTaxonomyDirectoryFormat to taxonomy.qza

• At this point, if you wish, you can average observations across samples according to sample metadata using groupsamples.sh.

  bash ../../../fast-16s-analysis/shell_scripts/groupsamples.sh Luminal_Colon_notax_Trios_ASV_table_Silva_v138_1.qza SLC_TOTAL_OCT2020_FULL_Metadata.tsv Genotype taxonomy.qza 

• If you grouped your observations, you will need to make a "metadata" .tsv file that denotes the levels of the factor by which you grouped your samples. Since I grouped by "Genotype" above, I will now make a "metadata" file that contains a single column: #SampleID WT HET MUT
• Now, I can use taxa-barplots.sh to make my barplots.

bash ../../../fast-16s-analysis/shell_scripts/taxabarplot.sh groupby_Genotype_Luminal_Colon_notax_Trios_ASV_table_Silva_v138_1.qza taxonomy.qza Genotype_Metadata.tsv