mix_gc_config.json

{"work_dir": "/abs/path/to/mix_gc_pipeline",
"path_to_ind_assembly_list": "Ind_contigs_list.csv",
"path_to_co_assembly_contigs": "/abs/path/to/co_assembly_contigs.fa.gz",
"prodigal_params": "-p meta -q",
"cluster_params":"-c 0.95 --min-seq-id 0.95 --cov-mode 1 --cluster-mode 2",
"list_specific_genes_to_remove": "None",
"tmp_dir":"TMP_dir",
"path_to_eggnog_db": "/abs/path/to/eggnog_db/data",
"path_dbcan_db":"dbcan_db/dbCAN-fam-HMMs.txt",
"path_pfam_db":"Pfam_db/Pfam-A.hmm",
"path_to_CAT_gtdb_dir": "/abs/path/to/CAT_GTDB_database",
"include_CAT_GTDB_taxonomy": True,
"n_chuncks_annotations": 20,
"n_chuncks_taxonomy": 50,
"taxonomy_DB": "uniprot",
"GTDB": {
        "GTDB_dir":"GTDB_aa_db/protein_faa_reps",
        "bact_tsv": "GTDB_aa_db/bac120_taxonomy.tsv",
        "arch_tsv": "GTDB_aa_db/ar53_taxonomy.tsv"
        },
"uniprot": {
            "uniprot_db_type":"UniRef90",
            "by_chuncks": False,
            },
"mmseqs_taxonomy_params":"--tax-lineage 1 -v 1 --report-mode 1 --min-seq-id 0.7",
"extra_mmseqs_taxonomy_params_Virus": "--orf-filter 1",
"performance": { "Big_jobs": {"Threads": 20,
                              "memory_per_cpu": 6400
                              },
                  "Small_jobs": {"Threads": 1,
                                 "memory_per_cpu": 6400
                                 },
                  "slurm_extra":"",
                },
}