AllTheBacteria/All-representative-ids-QC

List of sample identifiers from ATB, GTDB, MGnify, SPIRE, mOTUs and HRGM datasets, deduplicated at varying levels of identity based on nucleotide sequence using sketchlib v0.2.5.

Sketching and distance calculation conducted with below commands

sketchlib inverted build -o ${outpref} -k 21 -s 10 -f ${infile} --threads 47 --write-skq
sketchlib sketch -f ${infile} -o ${outpref} --k-vals 21 -s 1000 --threads 47

sketchlib inverted precluster --knn 50 --skd ${outpref} --ani --threads 47 -o ${outpref}_dists.tsv ${outpref}.ski

Where infile is a set of genome inputs in the format:

GENOME000001	/path/to/GENOME000001.fasta
GENOME000002	/path/to/GENOME000002.fasta
GENOME000003	/path/to/GENOME000003.fasta

Representatives were then generated using the get_representatives.py script:

python get_representatives.py --infile ${outpref}_dists.tsv --min-return 1 --workflow reps_graph --to-keep qc_list.csv --threshold 0.80,0.85,0.90,0.95,0.98,0.99,1.0 --outpref ${outpref} --workflow reps

Where qc_list.csv is set of sample ids (one per line) which passed QC measures:

• CheckM2 completeness of >=50%
• CheckM2 contamination of < 10%
• Quality score of >=50, calculated as completeness minus five times contamination
• <= 1,000 contigs
• Between 250 kb and 20 Mb in length
• Between 500 and 1,500 proteins/Mb