Scripts to reproduce the results of Buchfink et al., "Sensitive clustering of protein sequences at tree-of-life scale using DIAMOND DeepClust".
Data files that are required (see Data availability):
arch80_all.tsv: File mapping NCBI accessions to Pfam domain architectures.nr_accessions.tsv.zst: List of all accessions in NR database that was used for the main benchmark.reps_1M_ge5.tsv.zst: List of accessions of randomly sampled representatives of the big ~19bn clustering run of clusters of size >= 5.clan2acc.tsv: File mapping Pfam accessions to clans.reps_10k.ge2.faa: Sequences of 10k randomly sampled representatives of the big ~19bn clustering run of clusters of size >= 2.reps_10k.ge3.faa: Sequences of 10k randomly sampled representatives of the big ~19bn clustering run of clusters of size >= 3.clust.dedup2.tsv.zst: TSV file mapping cluster member sequences to representatives for the big ~19bn clustering run.centroids.dedup.faa: FASTA file of representatives for the big ~19bn clustering run.