config.yaml

#-- email notifications of pipeline success/failure (use "Skip" to deactivate) --#
email: None

#-- I/O --#
# file listing samples and associated data
samples_file: data/GTDBr95_n10/GTDBr95_n10.tsv

## column names in samples table
samples_col: 'ncbi_organism_name'
accession_col: 'accession'
fasta_file_path_col: 'fasta_file_path'
taxID_col: 'gtdb_taxid'          # or 'ncbi_species_taxid'
taxonomy_col: 'gtdb_taxonomy'    # or 'ncbi_taxonomy' 

# output location
output_dir: tests/output/GTDBr95_n10/

# temporary file directory (your username will be added automatically)
tmp_dir: /ebio/abt3_scratch/

#-- databases to create --#
# Replace "Create" with "Skip" to skip creation of any of these
# Note that braken relies on the kraken2 database
databases:
  kraken2: Create
  bracken: Create
  genes: Create
  humann3_bowtie2: Create
  humann3_diamond: Create

# Name of UniRef clustering (uniref90 or uniref50)
## "uniref90" highly recommended
uniref_name: uniref90
# Name of the humann3 diamond database to be created
## This must match the naming allowed by humann3 (eg., "uniref90_201901.dmnd")
dmnd_name: uniref90_201901.dmnd
# Index mapping UniRef90 clusters to UniRef50 (saves time vs re-annotating)
## This is skipped if annotating with UniRef50 instead of UniRef90
cluster_idx: data/uniref50-90.pkl

#-- if custom NCBI/GTDB taxdump files, "Skip" if standard NCBI taxdump --#
# Used for kraken taxonomy & metaphlan
names_dmp: data/taxdump/names.dmp
nodes_dmp: data/taxdump/nodes.dmp

#-- keep intermediate files required for re-creating DBs (eg., w/ more genomes) --#
# If "True", the intermediate files are saved to `output_dir`
# Else, the intermediate files are temporarily stored in `temp_folder`
keep_intermediate: True

#-- software parameters --#
# `vsearch_per_genome` = per-genome gene clustering
# use "Skip" at the start of any param to skip (if possible to skip) 
# for humann3, use either mmseqs or diamond (mmseqs gets priority if neither skipped)
# for humann3::mmseqs_search::run, --num-iterations must be >=2
params:
  ionice: -c 3
  bracken:
    build_kmer: 35
    build_read_lens:
      - 100
      - 150
  genes:
    prodigal: ""
    vsearch_per_genome: --id 0.97 --strand both --qmask none --fasta_width 0
    mmseqs_cluster: --min-seq-id 0.9 -c 0.8
    mmseqs_cluster_method: linclust     # or "cluster", which is slower
  humann3:
    batches: 2
    mmseqs_search:
      db: data/UniRef90/uniref90
      index: -s 6
      run: -e 1e-3 --max-accept 1 --max-seqs 100 --num-iterations 2 --start-sens 1 --sens-steps 3 -s 6
    diamond:
      db: Skip #data/uniref90_ec-filtered/uniref90_ec_filt_201901.dmnd
      run: --evalue 1e-3 --query-cover 80 --id 90 --max-target-seqs 1 --block-size 4 --index-chunks 2
    propagate_annotations: --min-cov 80 --min-pident 90

#-- snakemake pipeline --#
pipeline:
  snakemake_folder: ./
  script_folder: ./bin/scripts/
  name: Struo2_db-create
  config: create