Merge pull request #117 from nf-core/support-custom-panels

nf-core · Dec 5, 2024 · fc1bbf0 · fc1bbf0
2 parents f029bec + 786ce02
commit fc1bbf0
Show file tree

Hide file tree

Showing 13 changed files with 221 additions and 63 deletions.
diff --git a/.nf-core.yml b/.nf-core.yml
@@ -17,6 +17,10 @@ lint:
     - lib/Utils.groovy
     - lib/WorkflowMain.groovy
     - lib/WorkflowOncoanalyser.groovy
+  nextflow_config:
+    - config_defaults:
+        - params.fastp_umi_length
+        - params.fastp_umi_skip
 bump_version: null
 org_path: null
 update: null
diff --git a/lib/Utils.groovy b/lib/Utils.groovy
@@ -277,33 +277,26 @@ class Utils {
             // Apply some required restrictions to targeted mode
             if (run_config.mode === Constants.RunMode.TARGETED) {
 
-                // Do not allow normal DNA
-                if (Utils.hasNormalDna(meta)) {
-                    log.error "targeted mode is not compatible with the normal DNA BAM provided for ${meta.group_id}\n\n" +
-                        "The targeted workflow supports only tumor DNA BAMs (and tumor RNA BAMs for TSO500)"
+                // Do not allow donor DNA
+                if (Utils.hasDonorDna(meta)) {
+                    log.error "targeted mode is not compatible with the donor DNA BAM provided for ${meta.group_id}\n\n" +
+                        "The targeted workflow supports only tumor and normal DNA BAMs (and tumor RNA BAMs for TSO500)"
                     Nextflow.exit(1)
                 }
 
                 // Do not allow only tumor RNA
-                if (Utils.hasTumorRnaBam(meta) && !Utils.hasTumorDna(meta)) {
+                if (Utils.hasTumorRna(meta) && !Utils.hasTumorDna(meta)) {
                     log.error "targeted mode is not compatible with only tumor RNA provided for ${meta.group_id}\n\n" +
                         "The targeted workflow requires tumor DNA and can optionally take tumor RNA, depending on " +
                         "the configured panel."
                     Nextflow.exit(1)
                 }
 
-                // Restrict tumor RNA inputs to the TSO500 panel
-                if (Utils.hasTumorRnaBam(meta) && run_config.panel != 'tso500') {
-                    def panel = run_config.panel.toUpperCase()
-                    log.error "only the TSO500 panel supports tumor RNA analysis but got: ${panel}"
-                    Nextflow.exit(1)
-                }
-
             }
 
             // Do not allow normal DNA only
             if (Utils.hasNormalDna(meta) && !Utils.hasTumorDna(meta)) {
-                log.error "germline only mode not supported, found only a normal DNA BAM for ${meta.group_id}"
+                log.error "found only normal DNA input for ${meta.group_id} but germline only analysis is not supported"
                 Nextflow.exit(1)
             }
 

diff --git a/lib/WorkflowMain.groovy b/lib/WorkflowMain.groovy
@@ -37,7 +37,7 @@ class WorkflowMain {
             }
         }
 
-        if (!params.containsKey('ref_hmf_data_path')) {
+        if (!params.containsKey('ref_data_hmf_data_path')) {
             if (params.genome_version.toString() == '37') {
                 params.ref_data_hmf_data_path = Constants.HMF_DATA_37_PATH
             } else if (params.genome_version.toString() == '38') {
@@ -65,14 +65,30 @@ class WorkflowMain {
         if (run_mode === Constants.RunMode.TARGETED) {
 
             // Attempt to set default panel data path; make no assumption on valid 'panel' value
-
             if (params.containsKey('panel')) {
                 if (params.panel == 'tso500' && params.genome_version.toString() == '37') {
                     params.ref_data_panel_data_path = Constants.TSO500_PANEL_37_PATH
                 } else if (params.panel == 'tso500' && params.genome_version.toString() == '38') {
                     params.ref_data_panel_data_path = Constants.TSO500_PANEL_38_PATH
                 }
             }
+
+            // When fastp UMI is enabled, MarkDups UMI should be as well
+            if (params.fastp_umi && (!params.containsKey('markdups_umi') || !params.markdups_umi)) {
+                params.markdups_umi = true
+            }
+
+            // Set the MarkDups UMI duplex delimiter to '_' when the following conditions are met:
+            //   - both fastp and MarkDups UMI processing enabled
+            //   - fastp is using a duplex UMI location type (per_index or per_read)
+            //   - no MarkDups duplex delimiter has been set
+            def fastp_and_markdups_umi = params.fastp_umi && params.markdups_umi
+            def fastp_duplex_location = params.containsKey('fastp_umi_location') && (params.fastp_umi_location == 'per_index' || params.fastp_umi_location == 'per_read')
+            def no_umi_duplex_delim = !params.containsKey('markdups_umi_duplex_delim') || !params.markdups_umi_duplex_delim
+            if (fastp_and_markdups_umi && fastp_duplex_location && no_umi_duplex_delim) {
+                params.markdups_umi_duplex_delim = '_'
+            }
+
         }
 
         def stages = Processes.getRunStages(
@@ -93,12 +109,18 @@ class WorkflowMain {
         }
 
         // Final point to set any default to avoid access to undefined parameters during nf-validation
-        if (!params.containsKey('panel')) { params.panel = null }
-        if (!params.containsKey('ref_data_genome_alt')) { params.ref_data_genome_alt = null }
-        if (!params.containsKey('ref_data_genome_gtf')) { params.ref_data_genome_gtf = null }
-        if (!params.containsKey('ref_data_hla_slice_bed')) { params.ref_data_hla_slice_bed = null }
-        if (!params.containsKey('ref_data_panel_data_path')) { params.ref_data_panel_data_path = null }
-        if (!params.containsKey('ref_data_virusbreakenddb_path')) { params.ref_data_virusbreakenddb_path = null }
+        if (!params.containsKey('panel')) params.panel = null
+        if (!params.containsKey('ref_data_genome_alt')) params.ref_data_genome_alt = null
+        if (!params.containsKey('ref_data_genome_gtf')) params.ref_data_genome_gtf = null
+        if (!params.containsKey('ref_data_hla_slice_bed')) params.ref_data_hla_slice_bed = null
+        if (!params.containsKey('ref_data_panel_data_path')) params.ref_data_panel_data_path = null
+        if (!params.containsKey('ref_data_virusbreakenddb_path')) params.ref_data_virusbreakenddb_path = null
+
+        // Additionally set selected parameters with false-ish truthy values to avoid passing null values as inputs
+        if (!params.containsKey('fastp_umi_location')) params.fastp_umi_location = ''
+        if (!params.containsKey('fastp_umi_length')) params.fastp_umi_length = 0
+        if (!params.containsKey('fastp_umi_skip')) params.fastp_umi_skip = -1
+        if (!params.containsKey('markdups_umi_duplex_delim')) params.markdups_umi_duplex_delim = ''
 
     }
 
@@ -191,15 +213,20 @@ class WorkflowMain {
 
             } else if (!Constants.PANELS_DEFINED.contains(params.panel)) {
 
-                def panels = Constants.PANELS_DEFINED.join('\n    - ')
-                log.error "\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n" +
-                    "  The ${params.panel} is not defined. Currently, the available panels are:\n" +
-                    "    - ${panels}\n" +
-                    "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~"
-                Nextflow.exit(1)
+                if (params.containsKey('force_panel') && params.force_panel) {
+                    log.warn "provided panel ${params.panel} does not have built-in support but forcing to proceed"
+                } else {
+                    def panels = Constants.PANELS_DEFINED.join('\n    - ')
+                    log.error "\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n" +
+                        "  The ${params.panel} panel does not have built-in support. Currently, the\n" +
+                        "  available supported panels are:\n" +
+                        "    - ${panels}\n\n" +
+                        "  Please adjust the --panel argument or override with --force_panel.\n" +
+                        "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~"
+                    Nextflow.exit(1)
+                }
 
             }
-
         }
 
         if (params.ref_data_genome_alt !== null) {
@@ -219,8 +246,38 @@ class WorkflowMain {
                     "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~"
                 Nextflow.exit(1)
             }
+
         }
 
+        // UMI parameters
+
+        def fastp_umi_args_set_any = params.fastp_umi_location || params.fastp_umi_length || params.fastp_umi_skip >= 0
+        if (fastp_umi_args_set_any && !params.fastp_umi) {
+            log.error "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n" +
+                "  Detected use of fastp UMI parameters but fastp UMI processing has not been enabled.\n" +
+                "  Please review your configuration and set the fastp_umi flag or otherwise adjust\n" +
+                "  accordingly.\n" +
+                "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~"
+            Nextflow.exit(1)
+        }
+
+        def fastp_umi_args_set_all = params.fastp_umi_location && params.fastp_umi_length && params.fastp_umi_skip >= 0
+        if (params.fastp_umi && !fastp_umi_args_set_all) {
+            log.error "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n" +
+                "  Refusing to run fastp UMI processing without having any UMI params configured.\n" +
+                "  Please review your configuration and appropriately set all fastp_umi_* parameters.\n" +
+                "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~"
+            Nextflow.exit(1)
+        }
+
+        if (params.markdups_umi_duplex_delim && params.markdups_umi === false) {
+            log.error "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n" +
+                "  Detected use of MarkDups UMI parameters but MarkDups UMI processing has not been\n" +
+                "  enabled. Please review your configuration and set the markdups_umi flag or\n" +
+                "  otherwise adjust accordingly.\n" +
+                "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~"
+            Nextflow.exit(1)
+        }
 
     }
 

diff --git a/modules/local/fastp/main.nf b/modules/local/fastp/main.nf
@@ -10,6 +10,9 @@ process FASTP {
     input:
     tuple val(meta), path(reads_fwd), path(reads_rev)
     val max_fastq_records
+    val umi_location
+    val umi_length
+    val umi_skip
 
     output:
     tuple val(meta), path('*_R1.fastp.fastq.gz'), path('*_R2.fastp.fastq.gz'), emit: fastq
@@ -21,6 +24,14 @@ process FASTP {
     script:
     def args = task.ext.args ?: ''
 
+    def split_by_lines_arg = max_fastq_records > 0 ? "--split_by_lines ${4 * max_fastq_records}" : ''
+
+    def umi_args_list = []
+    if (umi_location) umi_args_list.add("--umi_loc ${umi_location}")
+    if (umi_length) umi_args_list.add("--umi_len ${umi_length}")
+    if (umi_skip >= 0) umi_args_list.add("--umi_skip ${umi_skip}")
+    def umi_args = umi_args_list ? '--umi ' + umi_args_list.join(' ') : ''
+
     """
     fastp \\
         ${args} \\
@@ -30,8 +41,8 @@ process FASTP {
         --disable_length_filtering \\
         --disable_adapter_trimming \\
         --disable_trim_poly_g \\
-        --split_by_lines ${4 * max_fastq_records} \\
-        --thread ${task.cpus} \\
+        ${umi_args} \\
+        ${split_by_lines_arg} \\
         --out1 ${meta.sample_id}_${meta.library_id}_${meta.lane}_R1.fastp.fastq.gz \\
         --out2 ${meta.sample_id}_${meta.library_id}_${meta.lane}_R2.fastp.fastq.gz
 

diff --git a/modules/local/fastp/meta.yml b/modules/local/fastp/meta.yml
@@ -27,7 +27,16 @@ input:
       pattern: "*.{fastq.gz}"
   - max_fastq_records:
       type: integer
-      description: Maximum number of reads per file
+      description: Maximum number of reads per file (optional)
+  - umi_location:
+      type: string
+      description: UMI location type (optional)
+  - umi_length:
+      type: integer
+      description: UMI length (optional)
+  - umi_skip:
+      type: integer
+      description: UMI base skip (optional)
 output:
   - meta:
       type: map

diff --git a/modules/local/markdups/main.nf b/modules/local/markdups/main.nf
@@ -14,7 +14,8 @@ process MARKDUPS {
     path genome_fai
     path genome_dict
     path unmap_regions
-    val has_umis
+    val umi_enable
+    val umi_duplex_delim
 
     output:
     tuple val(meta), path('*.markdups.bam'), path('*.markdups.bam.bai'), emit: bam
@@ -29,7 +30,12 @@ process MARKDUPS {
 
     def xmx_mod = task.ext.xmx_mod ?: 0.95
 
-    def umi_flags = has_umis ? '-umi_enabled -umi_duplex -umi_duplex_delim +' : ''
+    def form_consensus_arg = umi_enable ? '' : '-form_consensus'
+
+    def umi_args_list = []
+    if (umi_enable) umi_args_list.add('-umi_enabled')
+    if (umi_duplex_delim) umi_args_list.add("-umi_duplex -umi_duplex_delim ${umi_duplex_delim}")
+    def umi_args = umi_args_list ? umi_args_list.join(' ') : ''
 
     """
     markdups \\
@@ -42,8 +48,8 @@ process MARKDUPS {
         -sample ${meta.sample_id} \\
         -input_bam ${bams.join(',')} \\
         \\
-        -form_consensus \\
-        ${umi_flags} \\
+        ${form_consensus_arg} \\
+        ${umi_args} \\
         \\
         -unmap_regions ${unmap_regions} \\
         -ref_genome ${genome_fasta} \\
@@ -68,7 +74,7 @@ process MARKDUPS {
     touch ${meta.sample_id}.markdups.bam.bai
     touch ${meta.sample_id}.duplicate_freq.tsv
 
-    if [[ -n "${has_umis}" ]]; then
+    if [[ -n "${umi_enable}" ]]; then
         touch ${meta.sample_id}.umi_coord_freq.tsv
         touch ${meta.sample_id}.umi_edit_distance.tsv
         touch ${meta.sample_id}.umi_nucleotide_freq.tsv

diff --git a/modules/local/markdups/meta.yml b/modules/local/markdups/meta.yml
@@ -40,9 +40,12 @@ input:
       type: file
       description: Unmapped regions file
       pattern: "*.{tsv}"
-  - has_umis:
+  - umi_enable:
       type: boolean
-      description: Flag indicating presence of UMIs in reads
+      description: Flag to enable UMI processing
+  - umi_duplex_delim:
+      type: string
+      description: UMI duplex delimiter
 output:
   - meta:
       type: map

diff --git a/nextflow.config b/nextflow.config
@@ -15,21 +15,27 @@ params {
     // Workflow mode
     mode = null
 
+    // Force options
+    force_genome = false
+    force_panel  = false
+
+    // Read processing and alignment options
+    max_fastq_records = 10000000
+    fastp_umi         = false
+    markdups_umi      = false
+
     // Process configuration
     processes_manual  = false
     processes_include = null
     processes_exclude = null
 
     // Reference genome information; iGenomes is effectively disabled but retained for linting
     genome           = null
-    force_genome     = false
     igenomes_base    = 's3://ngi-igenomes/igenomes/'
     igenomes_ignore  = true
     hmf_genomes_base = 'https://pub-cf6ba01919994c3cbd354659947f74d8.r2.dev/genomes'
 
     // Other reference data and config exposed to user on CLI
-    max_fastq_records        = 10000000
-
     isofox_counts            = null
     isofox_gc_ratios         = null
     isofox_gene_ids          = null
@@ -266,7 +272,7 @@ plugins {
 validation {
     // NOTE(SW): entries here are generally have conditional defaults or are for internal use only
     defaultIgnoreParams = [
-        "genomes",
+        'genomes',
         'hmf_data_paths',
         'panel_data_paths',
         'ref_data_genome_fasta',
@@ -287,6 +293,10 @@ validation {
         'ref_data_hmf_data_path',
         'ref_data_panel_data_path',
         'ref_data_virusbreakenddb_path',
+        'fastp_umi_length',
+        'fastp_umi_location',
+        'fastp_umi_skip',
+        'markdups_umi_duplex_delim',
     ]
 
     lenientMode = true