nextflow.config

params {
    /* Input and output */
    genomes = "data/*.fa"                    // input genomes (FASTA format)
    outdir = "./results"                     // output directory
    
    /* Quality assessment */
    busco_db = 'none'                        // BUSCO db path for offline mode. 'none': download from Internet
    lineage = 'auto-euk'                     // 'auto', 'auto-prok', 'auto-euk', a dataset name (e.g 'fungi' or 
                                             // 'fungi_odb10') or a path (e.g. '/home/user/fungi_odb10')
    
    /* Genomes filtering */
    skip_filtering = false                   // skip genome filtering
    min_completeness = 50                    // discard sequences with less than 50% completeness
    max_contamination = 10                   // discard sequences with more than 10% contamination

    /* Dereplication */
    skip_dereplication = false               // skip the dereplication step
    ani_thr = 0.99                           // ANI threshold for dereplication (> 0.90)
    min_overlap = 0.30                       // minimum aligned fraction

    /* BUSCO SCG MSA */
    /* '--lineage' must be different from 'auto', 'auto-prok' and 'auto-euk' */
    skip_msa = false                         // skip BUSCO SCG MSA
   
    /* Phylogenetic tree inference */
    /* '--lineage' must be different from 'auto', 'auto-prok' and 'auto-euk' */
    /* '--skip_msa' must be 'false' */
    skip_tree = false                        // skip phylogenetic tree inference
    max_ncols = 5000                         // maximum number of MSA columns (taken randomly) for tree inference
    seed_cols = 42                           // random seed
    raxml_mode = "default"                   // RAxML mode , "default": default RAxML tree search algorithm or
                                             // "rbs": rapid bootstrapping full analysis
    raxml_nsearch = 10                       // "default" mode only: number of inferences on the original alignment using
                                             // distinct randomized MP trees (if ­10 is specified, RAxML will compute 
                                             // 10 distinct ML trees starting from 10 distinct randomized maximum parsimony
                                             // starting trees)
    raxml_nboot = "autoMRE"                  // "rbs" mode only: bootstrap convergence criterion or number of bootstrap 
                                             // searches (see -I and -#/-N options in RAxML)

    

    /* Gene prediction */
    skip_genepred = false                    // skip gene prediction using MetaEuk (see also the mmseqs_db parameter)

    /* Taxonomic classification */
    skip_taxonomy = false                    // skip taxonomy classification using MMseq2 (see also the mmseqs_db parameter)

    /* MMseq2 database for gene prediction and taxonomic classification */
    mmseqs_db = "none"                       // MMseqs2 database path. 'none': download from Internet
                                             // see the mmseqs_db_name parameter.
    mmseqs_db_name = "UniProtKB/Swiss-Prot"  // used if if mmseqs_db_path = 'none'. The available
                                             // databases are listed at 
                                             // https://github.com/soedinglab/mmseqs2/wiki#downloading-databases.
   
    /* eggNOG options */
    skip_eggnog = false                      // skip eggNOG annotation
    eggnog_db = "none"                       // eggNOG v5.0 db dir. 'none': download from Internet
    eggnog_db_mem = false                    // store the eggNOG sqlite DB into memory (~44GB memory
                                             // required). This increase the annotation speed.

    /* ITSx options */
    skip_itsx = false                        // skip ITSx extraction (SSU, ITS1, 5.8S, ITS2 and LSU)
    
    /* Limits */
    max_cpus = 8
    max_memory = 240.GB
    max_time = 240.h
}    

if( params.eggnog_db_mem ) {
    eggnog_mapper_startmem = 48.GB
    }
else {
    eggnog_mapper_startmem = 12.GB
    }

/* Docker options */
docker.enabled = true
docker.runOptions = '-u \$(id -u):\$(id -g)'


/* Import process configuration file*/
includeConfig 'process.config'


/* Manifest */
manifest {
    homePage = 'metashot.github.io'
    description = 'An Automated Workflow for Eukaryotic MAGs'
    mainScript = 'main.nf'
    version = '1.1.0'
}


/* Functions */
def check_max(obj, max) {
    // see https://github.com/nextflow-io/nextflow/issues/640
    if( obj instanceof nextflow.util.MemoryUnit ) {
        try {
            def max_type = max as nextflow.util.MemoryUnit
            return obj.compareTo(max_type) == 1 ? max_type : obj
        }   
        catch( all ) {
            println "ERROR: invalid max memory '${max}', using default value: $obj"
            return obj
        }   
    }
    if( obj instanceof nextflow.util.Duration ) {
        try {
            def max_type = max as nextflow.util.Duration
            return obj.compareTo(max_type) == 1 ? max_type : obj
        }   
        catch( all ) {
            println "ERROR: invalid max time '${max}', using default value $obj"
            return obj
        }         
    }
    if( obj instanceof Integer ) {
        try {
            return Math.min(obj, max as int)
        }   
        catch( all ) {
            println "ERROR: invalid max cpus '${max}', using default value $obj"
            return obj
        }         
    }    
}