fixed quantms-rescoring

modules/local/utils/msrescore_features/main.nf

@@ -3,17 +3,17 @@ process MSRESCORE_FEATURES {
     label 'process_high'
 
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'oras://ghcr.io/bigbio/quantms-rescoring-sif:0.0.13' :
-        'ghcr.io/bigbio/quantms-rescoring:0.0.13' }"
+        'oras://ghcr.io/bigbio/quantms-rescoring-sif:0.0.14' :
+        'ghcr.io/bigbio/quantms-rescoring:0.0.14' }"
 
     input:
-    tuple val(meta), path(idxml), path(mzml), path(model_weight)
+    tuple val(meta), path(idxml), path(mzml), path(model_weight), val(search_engine)
 
     output:
-    tuple val(meta), path("*ms2rescore.idXML") , emit: idxml
-    tuple val(meta), path("*.html" )           , optional:true, emit: html
-    path "versions.yml"                        , emit: versions
-    path "*.log"                               , emit: log
+    tuple val(meta), path("*ms2rescore.idXML"), val(search_engine) , emit: idxml
+    tuple val(meta), path("*.html" )                               , optional:true, emit: html
+    path "versions.yml"                                            , emit: versions
+    path "*.log"                                                   , emit: log
 
     when:
     task.ext.when == null || task.ext.when
@@ -27,8 +27,10 @@ process MSRESCORE_FEATURES {
     // When --ms2features_model_dir is passed with no value, Nextflow may set it to boolean true
     if (params.ms2features_fine_tuning) {
         ms2_model_dir = '--ms2_model_dir ./'
-    } else {
+    } else if (params.ms2features_model_dir && params.ms2features_model_dir != true){
         ms2_model_dir = "--ms2_model_dir ${model_weight}"
+    } else {
+        ms2_model_dir = "--ms2_model_dir ./"
     }
 
     // Determine if using ms2pip or alphapeptdeep based on ms2features_generators

modules/local/utils/msrescore_fine_tuning/main.nf

@@ -3,31 +3,37 @@ process MSRESCORE_FINE_TUNING {
     label 'process_high'
 
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'oras://ghcr.io/bigbio/quantms-rescoring-sif:0.0.13' :
-        'ghcr.io/bigbio/quantms-rescoring:0.0.13' }"
+        'oras://ghcr.io/bigbio/quantms-rescoring-sif:0.0.14' :
+        'ghcr.io/bigbio/quantms-rescoring:0.0.14' }"
 
     input:
-    tuple val(meta), path(idxml), path(mzml), path(ms2_model_dir)
+    tuple val(meta), path(idxml), path(mzml), val(groupkey), path(ms2_model_dir)
 
     output:
-    path "retained_ms2.pth" , emit: model_weight
-    path "versions.yml"     , emit: versions
-    path "*.log"            , emit: log
+    tuple val(groupkey), path("retained_ms2.pth") , emit: model_weight
+    path "versions.yml"                           , emit: versions
+    path "*.log"                                  , emit: log
 
     when:
     task.ext.when == null || task.ext.when
 
     script:
     def args = task.ext.args ?: ''
-    def prefix = task.ext.prefix ?: "${meta.mzml_id}_ms2rescore"
+    def prefix = task.ext.prefix ?: "${groupkey}_fine_tuning"
 
     // Initialize tolerance variables
     def ms2_tolerance = null
     def ms2_tolerance_unit = null
 
     // ms2pip only supports Da unit, but alphapeptdeep supports both Da and ppm
-    ms2_tolerance = meta['fragmentmasstolerance']
-    ms2_tolerance_unit = meta['fragmentmasstoleranceunit']
+    ms2_tolerance = meta[0]['fragmentmasstolerance']
+    ms2_tolerance_unit = meta[0]['fragmentmasstoleranceunit']
+
+    if (params.ms2features_model_dir && params.ms2features_model_dir != true) {
+        ms2_model_dir = ms2_model_dir
+    } else {
+        ms2_model_dir = "./"
+    }
 
     if (params.force_transfer_learning) {
         force_transfer_learning = "--force_transfer_learning"
@@ -56,7 +62,7 @@ process MSRESCORE_FINE_TUNING {
         ${force_transfer_learning} \\
         ${consider_modloss} \\
         $args \\
-        2>&1 | tee ${idxml.baseName}_fine_tuning.log
+        2>&1 | tee ${groupkey}_fine_tuning.log
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

modules/local/utils/spectrum_features/main.nf

@@ -7,10 +7,10 @@ process SPECTRUM_FEATURES {
         'ghcr.io/bigbio/quantms-rescoring:0.0.13' }"
 
     input:
-    tuple val(meta), path(id_file), path(ms_file)
+    tuple val(meta), path(id_file), val(search_engine), path(ms_file)
 
     output:
-    tuple val(meta), path("${id_file.baseName}_snr.idXML"), emit: id_files_snr
+    tuple val(meta), path("${id_file.baseName}_snr.idXML"), val(search_engine), emit: id_files_snr
     path "versions.yml", emit: versions
     path "*.log", emit: log
 

subworkflows/local/dda_id/main.nf

@@ -3,14 +3,8 @@
 //
 include { CONSENSUSID          } from '../../../modules/local/openms/consensusid/main'
 include { PERCOLATOR           } from '../../../modules/local/openms/percolator/main'
-include { ID_MERGER            } from '../../../modules/local/openms/id_merger/main'
 include { ID_RIPPER            } from '../../../modules/local/openms/id_ripper/main'
 include { PSM_CONVERSION       } from '../../../modules/local/utils/psm_conversion/main'
-include { MSRESCORE_FEATURES   } from '../../../modules/local/utils/msrescore_features/main'
-include { GET_SAMPLE           } from '../../../modules/local/utils/extract_sample/main'
-include { SPECTRUM_FEATURES    } from '../../../modules/local/utils/spectrum_features/main'
-include { PSM_CLEAN            } from '../../../modules/local/utils/psm_clean/main'
-include { MSRESCORE_FINE_TUNING} from '../../../modules/local/utils/msrescore_fine_tuning/main'
 include { PHOSPHO_SCORING      } from '../phospho_scoring/main'
 
 //
@@ -35,10 +29,11 @@ workflow DDA_ID {
     //
     PEPTIDE_DATABASE_SEARCH (
         ch_file_preparation_results,
-        ch_database_wdecoy
+        ch_database_wdecoy,
+        ch_expdesign
     )
     ch_software_versions = ch_software_versions.mix(PEPTIDE_DATABASE_SEARCH.out.versions)
-    ch_id_files = PEPTIDE_DATABASE_SEARCH.out.ch_id_files_idx
+    ch_id_files_feats = PEPTIDE_DATABASE_SEARCH.out.ch_id_files_idx
 
     ch_pmultiqc_consensus = Channel.empty()
     ch_pmultiqc_ids = Channel.empty()
@@ -47,116 +42,14 @@ workflow DDA_ID {
     // SUBWORKFLOW: Rescoring
     //
     if (params.skip_rescoring == false) {
-
-        if (params.ms2features_enable == true) {
-
-            // Only add ms2_model_dir if it's actually set and not empty
-            // Handle cases where parameter might be empty string, null, boolean true, or whitespace
-            // When --ms2features_model_dir is passed with no value, Nextflow may set it to boolean true
-            if (params.ms2features_model_dir && params.ms2features_model_dir != true) {
-                ms2_model_dir = Channel.from(file(params.ms2features_model_dir, checkIfExists: true))
-            } else {
-                ms2_model_dir = Channel.from(file("./"))
-            }
-
-            if (params.ms2features_fine_tuning == true) {
-                if (params.ms2features_generators.toLowerCase().contains('ms2pip')) {
-                    exit(1, 'Error: Fine tuning only supports AlphaPeptdeep!')
-                } else {
-                    train_datasets = ch_id_files.combine(ch_file_preparation_results, by: 0).randomSample(params.fine_tuning_sample_run, 2025)
-                    MSRESCORE_FINE_TUNING(train_datasets.collect().combine(ms2_model_dir))
-                    MSRESCORE_FEATURES(ch_id_files.combine(ch_file_preparation_results, by: 0).combine(MSRESCORE_FINE_TUNING.out.model_weight))
-                    ch_software_versions = ch_software_versions.mix(MSRESCORE_FEATURES.out.versions)
-                    ch_id_files_feats = MSRESCORE_FEATURES.out.idxml
-                }
-            } else{
-                MSRESCORE_FEATURES(ch_id_files.combine(ch_file_preparation_results, by: 0).combine(ms2_model_dir))
-                ch_software_versions = ch_software_versions.mix(MSRESCORE_FEATURES.out.versions)
-                ch_id_files_feats = MSRESCORE_FEATURES.out.idxml
-            }
-
-        } else {
-            PSM_CLEAN(ch_id_files.combine(ch_file_preparation_results, by: 0))
-            ch_id_files_feats = PSM_CLEAN.out.idxml
-            ch_software_versions = ch_software_versions.mix(PSM_CLEAN.out.versions)
-        }
-
-        // Add SNR features to percolator
-        if (params.ms2features_snr) {
-            SPECTRUM_FEATURES(ch_id_files_feats.combine(ch_file_preparation_results, by: 0))
-            ch_id_files_feats = SPECTRUM_FEATURES.out.id_files_snr
-            ch_software_versions = ch_software_versions.mix(SPECTRUM_FEATURES.out.versions)
-        }
-
         // Rescoring for independent run, Sample or whole experiments
         if (params.ms2features_range == "independent_run") {
             PERCOLATOR(ch_id_files_feats)
             ch_software_versions = ch_software_versions.mix(PERCOLATOR.out.versions)
             ch_consensus_input = PERCOLATOR.out.id_files_perc
-        } else if (params.ms2features_range == "by_sample") {
-            // Sample map
-            GET_SAMPLE(ch_expdesign)
-            ch_software_versions = ch_software_versions.mix(GET_SAMPLE.out.versions)
-
-            ch_expdesign_sample = GET_SAMPLE.out.ch_expdesign_sample
-            ch_expdesign_sample.splitCsv(header: true, sep: '\t')
-                .map { get_sample_map(it) }.set{ sample_map_idv }
-
-            ch_id_files_feats.map {[it[0].mzml_id, it[0], it[1]]}
-                .combine(sample_map_idv, by: 0)
-                .map {[it[1], it[2], it[3]]}
-                .set{ch_id_files_feats_sample}
-
-            // Group by search_engines and sample
-            ch_id_files_feats_sample.branch{ meta, filename, sample  ->
-                sage: filename.name.contains('sage')
-                    return [meta, filename, sample]
-                msgf: filename.name.contains('msgf')
-                    return [meta, filename, sample]
-                comet: filename.name.contains('comet')
-                    return [meta, filename, sample]
-            }.set{ch_id_files_feat_branched}
-
-            // IDMERGER for samples group
-            ID_MERGER(ch_id_files_feat_branched.comet.groupTuple(by: 2)
-                .mix(ch_id_files_feat_branched.msgf.groupTuple(by: 2))
-                .mix(ch_id_files_feat_branched.sage.groupTuple(by: 2)))
-            ch_software_versions = ch_software_versions.mix(ID_MERGER.out.versions)
-
-            PERCOLATOR(ID_MERGER.out.id_merged)
-            ch_software_versions = ch_software_versions.mix(PERCOLATOR.out.versions)
-
-            // Currently only ID runs on exactly one mzML file are supported in CONSENSUSID. Split idXML by runs
-            ID_RIPPER(PERCOLATOR.out.id_files_perc)
-            ch_file_preparation_results.map{[it[0].mzml_id, it[0]]}.set{meta}
-            ID_RIPPER.out.id_rippers.flatten().map { add_file_prefix (it)}.set{id_rippers}
-            meta.combine(id_rippers, by: 0)
-                    .map{ [it[1], it[2]]}
-                    .set{ ch_consensus_input }
-            ch_software_versions = ch_software_versions.mix(ID_RIPPER.out.versions)
-
-        } else if (params.ms2features_range == "by_project"){
-            ch_id_files_feats.map {[it[0].experiment_id, it[0], it[1]]}.set { ch_id_files_feats}
-
-            // Split ch_id_files_feats by search_engines
-            ch_id_files_feats.branch{ experiment_id, meta, filename ->
-                sage: filename.name.contains('sage')
-                    return [meta, filename, experiment_id]
-                msgf: filename.name.contains('msgf')
-                    return [meta, filename, experiment_id]
-                comet: filename.name.contains('comet')
-                    return [meta, filename, experiment_id]
-            }.set{ch_id_files_feat_branched}
-
-            // IDMERGER for whole experiments
-            ID_MERGER(ch_id_files_feat_branched.comet.groupTuple(by: 2)
-                .mix(ch_id_files_feat_branched.msgf.groupTuple(by: 2))
-                .mix(ch_id_files_feat_branched.sage.groupTuple(by: 2)))
-            ch_software_versions = ch_software_versions.mix(ID_MERGER.out.versions)
-
-            PERCOLATOR(ID_MERGER.out.id_merged)
+        } else {
+            PERCOLATOR(ch_id_files_feats)
             ch_software_versions = ch_software_versions.mix(PERCOLATOR.out.versions)
-
             // Currently only ID runs on exactly one mzML file are supported in CONSENSUSID. Split idXML by runs
             ID_RIPPER(PERCOLATOR.out.id_files_perc)
             ch_file_preparation_results.map{[it[0].mzml_id, it[0]]}.set{meta}
@@ -165,7 +58,6 @@ workflow DDA_ID {
                     .map{ [it[1], it[2]]}
                     .set{ ch_consensus_input }
             ch_software_versions = ch_software_versions.mix(ID_RIPPER.out.versions)
-
         }
 
         ch_rescoring_results = ch_consensus_input

subworkflows/local/id/main.nf

@@ -26,7 +26,8 @@ workflow ID {
     //
     PEPTIDE_DATABASE_SEARCH (
         ch_file_preparation_results,
-        ch_database_wdecoy
+        ch_database_wdecoy,
+        ch_expdesign
     )
     ch_software_versions = ch_software_versions.mix(PEPTIDE_DATABASE_SEARCH.out.versions)