Code for tax-aliquots (Assigning Lineage to Queries Over Two Steps)

This code has been made available for testing and reproducibility purposes in conjunction with the manuscript "Missing microbial eukaryotes and misleading meta-omic conclusions". In the future, tax-aliquots will be incorporated into the EUKulele taxonomic annotation tool (https://github.com/AlexanderLabWHOI/EUKulele).

Arguments

• -i : input file from DeepClust (https://github.com/bbuchfink/diamond)
• -c : FASTA file that contains all sequences present in the DeepClust file, both reference and query
• --output_file : prefix for output file names
• --output_dir : directory in which to save output
• -t : taxonomy table, in the style of EUKulele
• --prot_map : protein name mapping, in the style of EUKulele
• --kmer_len : amino acid k-mer length to use
• -s : sample delimiter in sequence file
• --level_of_interest : taxonomic level of interest (e.g. phylum, family; should match EUKulele
• --list_of_interest : underscore-separated list of taxonomic labels to pull from the DeepClust file

Invocation

tax-aliquots can be invoked using the process_clusters.py file in src/tax-aliquots-scripts. tax-aliquots uses faSomeRecords (https://github.com/santiagosnchez/faSomeRecords) to read in records from fasta files

Example command

python process_clusters.py \
    -i deepclust_contigname.mad.50.out \
    -c combined_seqs.fasta \
    --output_file="test" \
    --output_dir="test_dir" \
    -t tax-table.txt \
    --prot_map=prot-map.json \
    --kmer_len=3 -s test --level_of_interest family \
    --list_of_interest Hemiaulaceae_Rhizosoleniaceae_Thalassiosiraceae_Skeletonemataceae

Citations

• Buchfink, Benjamin, et al. "Sensitive clustering of protein sequences at tree-of-life scale using DIAMOND DeepClust." bioRxiv (2023): 2023-01.
• Sanchez-Ramirez, Santiago, et al. "faSomeRecords." https://github.com/santiagosnchez/faSomeRecords