Skip to content

Latest commit

 

History

History
61 lines (46 loc) · 3.46 KB

Taxonomy.md

File metadata and controls

61 lines (46 loc) · 3.46 KB

Extracting Taxonomy Information Using Diamond Blast Report (IMG/VR Database)

Note: Run this command in the root folder where the SOVAP pipeline is executed.

This command processes all directories in the current folder. For each directory, it extracts data from output.diamond.tsv, joins it with taxonomy information from the IMGVR_all_Sequence_information.tsv file, and outputs a .taxo file with the extracted taxonomy data.

for folder in */; do \
    parent_dir=${folder%/}; \
    cut -f1,2 ${folder}6_Diamond-Taxonomy/output.diamond.tsv | cut -d "|" -f1 | \
        awk '{print $2 "\t" $1}' | \
        sort -k1,1 | \
        join -a1 -t $'\t' - /path/to/IMGVR_all_Sequence_information_sorted.tsv > \
        ${folder}6_Diamond-Taxonomy/$parent_dir.taxo; \
done

Details:

  1. The script loops through each subdirectory in the root folder.
  2. For each directory:
    • Extract relevant columns from output.diamond.tsv (column 1 and 2).
    • Clean the taxonomic information, removing everything after the first | symbol.
    • Format the output by swapping the order of columns, and sort the data by the first column.
    • Join the extracted data with the IMGVR_all_Sequence_information.tsv file, which contains sequence metadata. The -a1 option ensures that all lines from the first file (the blast output) are included, even if no match is found in the second file.
  3. Save the output to a file called foldername.taxo within the 6_Diamond-Taxonomy/ subdirectory of each folder.

The resulting .taxo file will contain taxonomic details for each sequence in the corresponding output.diamond.tsv file, ready for downstream analysis.

Important:

  • Ensure the IMGVR_all_Sequence_information.tsv file is pre-sorted by the first column (sequence ID) for the join operation to work correctly. Using an unsorted file may result in missing matches or errors. If the file is not sorted, you can pre-sort it as follows:
sort -k1,1 /path/to/IMGVR_all_Sequence_information.tsv -o /path/to/IMGVR_all_Sequence_information_sorted.tsv

Alternatively, if the file is not pre-sorted and sorting it is not feasible due to its size, you can sort the file during execution using the following command. However, this is not recommended when working with many directories, as it significantly increases runtime:

for folder in */; do \
    parent_dir=${folder%/}; \
    cut -f1,2 ${folder}6_Diamond-Taxonomy/output.diamond.tsv | cut -d "|" -f1 | \
        awk '{print $2 "\t" $1}' | \
        sort -k1,1 | \
        join -a1 -t $'\t' - <(sort -k1,1 /path/to/IMGVR_all_Sequence_information.tsv) > \
        ${folder}6_Diamond-Taxonomy/$parent_dir.taxo; \
done

Explanation:

  • In this command, the IMGVR_all_Sequence_information.tsv file is sorted on-the-fly during the join operation. This adds overhead, particularly when processing many directories, as the file is sorted every time the command is run.

Due to the large size of IMGVR_all_Sequence_information.tsv file, this step is provided as an optional side script. However, it is recommended to perform this step for reproducing graphs in the manuscript and analyzing data in R for taxonomy, diversity indices, and etc.

IMGVR_all_Sequence_information.tsv: a table listing the characteristics of each viral sequence such as its origin, affiliation, and predicted host (tsv format).

More information: https://genome.jgi.doe.gov/portal/IMG_VR/IMG_VR.home.html