nextflow.config

manifest {
  name = 'ScaleRna'
  version = '2.1.0'
  description = 'ScaleBio Seq Suite: RNA workflow'
  homePage = 'https://scale.bio'
}

includeConfig 'modules/create_mtx.config'
includeConfig 'modules/downstream.config'
includeConfig 'modules/scale_plex.config'
includeConfig 'modules/internal.config'

//// Parameter defaults; Can be set at workflow runtime on the nextflow command-line
// See nextflow_schema.json for description
params {
    //// Sample sheet, i.e. path to samples.csv (required!)
    samples = null

    //// Reference genome.json (required!)
    genome = null

    //// Sequencing data input.
    // For alignment, either a runfolder (BCL) or a directory with fastq files is required!
    runFolder = null
    fastqDir = null 
    fastqSamplesheet = null // Optional samplesheet for bcl-convert, replacing the auto-generated one
    ultimaCramDir = null // For QuantumScale data sequenced on Ultima
    // 'Reporting'-workflow, i.e. starting from existing alignment results
    reporting = false // Run only cell-filtering and reporting steps, with optional sample merging
    resultDir = null // Output directory for previous pipeline run (to combine multiple runs, specify in samples.csv)
    
    //// Output options
    outDir = "ScaleRna.out" // Workflow output directory
    bamOut = false // Set to false to skip BAM output from STAR
    bcParserBamOut = false // Set to true to publish unaligned BAMs from bcParser
    fastqOut = false // set to true to publish bcl converted fastq files to outDir

    //// Library structure (barcode locations and sequences).
    // Can be absolute paths or relative to ${projectDir}/references/
    // QuantumScale RNA: "libQuantumV1.0.json"
    // Scale RNA v1.1: "libV1.1.json"
    // Scale RNA v1.0: "libV1.json"
    libStructure = "libQuantumV1.0.json"
    
    // Enable ScalePlex integration
    scalePlex = false
    // 3L kit: "scaleplexlibV1.json"
    // QuantumScale kit: "scaleplexlibQuantumV1.0.json"
    scalePlexLibStructure = "scaleplexlibQuantumV1.0.json"
    // ScalePlex PCR to RNA PCR mapping for merging ScalePlex library with RNA library
    // Default mentioned in scaleplexlibQuantumV1.json, setting this param overrides the default
    scalePlexToRnaMapping = null

    //// Optional workflow parameters
    merge = true // Generate merged sample outputs across libraries / plates
    bclConvertParams = "" //"--bcl-only-lane 1"
    fastqc = true // Run fastqc on input fastqs

    starFeature = "GeneFull_Ex50pAS" // What read to transcript overlaps STAR counts (--soloFeatures)
    starMulti = "PropUnique" // How to handle reads matching multiple genes (--soloMultiMappers)
    starMultiBarnyard = "Unique" // How to handle reads matching multiple genes for barnyard datasets. (--soloMultiMappers)
    starStrand = "Forward" // Strandedness of RNA reads relative to annotated genes
    starMaxLoci = 6 // Maximum number of Loci multimapping reads can map (--outFilterMultimapNmax)
    roundCounts = false // Round UMI counts to nearest integer
    
    starTrimming = "" // Trimming in STAR (in addition to cutadapt)

    //// Cell Thresholding parameters
    // See modules/createMtx.config for more options
    minUTC = 100 // Minimum transcript count to consider a barcode as a potential cell
    cellFinder = true // Compare cell-barcode expression profiles to background to call more cells
    
    fixedCells = false // Call the top "expectedCells" many barcodes per sample as cells (set in samples.csv)
    UTC = 0 // Set a fixed threshold above which all barcodes are called (minUTC < X < UTC go to CellFinder if enabled)

    //// Resources and parallelization 
    splitFastq = true // Split fastqs by PCR and RT wells for increased parallelization
    // Max. resources that can be requested for a single task / process
    taskMaxMemory = 256.GB
    taskMaxCpus = 16
    taskMaxTime = 48.h
    // Number of RT barcodes per STAR job when using splitFastq
    rtBarcodesPerStarJob = 4
    // Number of bcParser jobs
    bcParserJobsPerLibName = 16
    // Number of total fastqc jobs
    totalFastqcJobs = 100
    // Number of files (input index2 fastq / input ucram) batched into a single LibraryDetection process
    filesPerLibDetection = 32
    
    //// Downstream Analysis
    // Enable preliminary seurat clustering and azimuth cell-type classification
    seurat = false
    azimuth = false
    compSamples = false
    // Enables output of UMI count matrices in annData format
    annData = false
    
    testRun = false // Disable some validations for small test-runs
    help = false
}

process {
    // Default errorStrategy is to retry if process exit code is 137, 138, 139 or 140 or if it is the first or second task attempt
    errorStrategy = { ((task.attempt) <= 2 && (task.exitStatus) != 42) || (task.exitStatus in 137..140) ? 'retry' : 'terminate' }
    maxRetries = 3

    cpus = { max_cpu(2) }
    memory = { max_mem(4.GB * task.attempt) }
    time = params.taskMaxTime

    container = "public.ecr.aws/o5l3p3e4/scalerna:2025-04-16-160428"

    withLabel:small {
        cpus = 1
        memory = { max_mem(2.GB * task.attempt) }
    }
    withLabel:large {
        cpus = { max_cpu(5) }
        memory = { max_mem(30.GB * task.attempt) }
    }
    withLabel:optional {
      errorStrategy = { (task.attempt) <= 2 || (task.exitStatus in 137..140) ? 'retry' : 'ignore' }
    }
    withName:BclConvert {
        container = 'nfcore/bclconvert:3.9.3'
        cpus = { max_cpu(16) }
        memory = { max_mem(120.GB * task.attempt) }
    }
    withName:TrimFq {
	    cpus = { max_cpu(8 * task.attempt) }
	    memory = { max_mem(8.GB * task.attempt) }
    }
    withName:BarcodeParser {
        cpus = { max_cpu(5) }
        memory = { max_mem(6.GB * task.attempt) }
    }
    withName:StarSolo {
        cpus = { max_cpu(16) }
        memory = { max_mem(60.GB * task.attempt) }
    }
    withLabel:report {
        container = "public.ecr.aws/o5l3p3e4/scalernareport:2025-04-16-161858"
        cpus = { max_cpu(5) }
        memory = { max_mem(30.GB * task.attempt) }
    }
}
profiles {
  conda {
    conda.enabled = true
    process.conda = "$projectDir/envs/scalerna.conda.yml"
    process {
      withLabel:report {
        conda = "$projectDir/envs/scalernareport.conda.yml"
      }
    }
  }
  docker {
    // Shared settings for all container engines go here
    docker.enabled = true
    docker.fixOwnership = true
    // docker.runOptions = '-u $(id -u):$(id -g)' // Alt. to fixOwnership; match user ID in container to outside
  }
  singularity {
    singularity.enabled = true 
    singularity.autoMounts = true
    docker.enabled = false
  }
  podman {
    podman.enabled = true 
    docker.enabled = false
  }
}
conda.createTimeout = '1 h'

// nf-core functions to ensure that resource requirements don't go 
// beyond a maximum limit
def max_mem(obj) {
    if (obj.compareTo(params.taskMaxMemory as nextflow.util.MemoryUnit) == 1)
        return params.taskMaxMemory as nextflow.util.MemoryUnit
    else
        return obj
}
def max_cpu(obj) {
    return Math.min(obj as int, params.taskMaxCpus as int)
}