Merge pull request #207 from nf-core/203-nfcore_circrnacircrnabsj_detectioncombinebeds_reads-attributeerror-index-object-has-no-attribute-levels

Fix failure in BSJ read agreement analysis, if only a single read-yielding tool is enabled

Author: nictru

modules/local/combinebeds/reads/main.nf

@@ -12,8 +12,8 @@ process COMBINEBEDS_READS {
 
     output:
     tuple val(meta), path("${prefix}.bed"), emit: combined
-    path "*.png"                          , emit: plots
-    path "*.json"                         , emit: multiqc
+    path "*.png"                          , emit: plots, optional: true
+    path "*.json"                         , emit: multiqc, optional: true
     path "versions.yml"                   , emit: versions
 
     script:

modules/local/combinebeds/reads/templates/reads.py

@@ -38,31 +38,32 @@
 # Investigate read-level agreement
 df_bsj_groups = df.group_by(*bsj_columns, "reads").agg(pl.col("tool").unique()).collect()
 
-# Plot read-level agreement
-bsj_upset_data = from_memberships(df_bsj_groups["tool"].to_list())
-bsj_upset = UpSet(bsj_upset_data, show_counts=True, subset_size="count")
-bsj_upset.plot()
-plot_file = "${prefix}:read_agreement.png"
-plt.savefig(plot_file)
-plt.close()
-
-# MultiQC
-image_string = base64.b64encode(open(plot_file, "rb").read()).decode("utf-8")
-image_html = f'<div class="mqc-custom-content-image"><img src="data:image/png;base64,{image_string}" /></div>'
-
-multiqc = {
-    'id': f"{meta_id}_read_agreement",
-    'parent_id': "read_agreement",
-    'parent_name': 'Read agreement',
-    'parent_description': 'While different tools may agree on the presence of a BSJ, they may disagree on the reads that support it. These plots show the read-level agreement between tools for each BSJ.',
-    'section_name': meta_id,
-    'description': f'UpSet plot showing the read-level agreement between tools for each BSJ in {meta_id}',
-    'plot_type': 'image',
-    'data': image_html
-}
+if len(tools) > 1:
+    # Plot read-level agreement
+    bsj_upset_data = from_memberships(df_bsj_groups["tool"].to_list())
+    bsj_upset = UpSet(bsj_upset_data, show_counts=True, subset_size="count")
+    bsj_upset.plot()
+    plot_file = "${prefix}:read_agreement.png"
+    plt.savefig(plot_file)
+    plt.close()
+
+    # MultiQC
+    image_string = base64.b64encode(open(plot_file, "rb").read()).decode("utf-8")
+    image_html = f'<div class="mqc-custom-content-image"><img src="data:image/png;base64,{image_string}" /></div>'
+
+    multiqc = {
+        'id': f"{meta_id}_read_agreement",
+        'parent_id': "read_agreement",
+        'parent_name': 'Read agreement',
+        'parent_description': 'While different tools may agree on the presence of a BSJ, they may disagree on the reads that support it. These plots show the read-level agreement between tools for each BSJ.',
+        'section_name': meta_id,
+        'description': f'UpSet plot showing the read-level agreement between tools for each BSJ in {meta_id}',
+        'plot_type': 'image',
+        'data': image_html
+    }
 
-with open(f"{meta_id}_read_agreement_mqc.json", "w") as f:
-    f.write(json.dumps(multiqc, indent=4))
+    with open(f"{meta_id}_read_agreement_mqc.json", "w") as f:
+        f.write(json.dumps(multiqc, indent=4))
 
 df_bsj_groups = df_bsj_groups.with_columns(
     tool_count = pl.col("tool").list.len()

subworkflows/local/combine_transcriptomes.nf

@@ -27,10 +27,6 @@ workflow COMBINE_TRANSCRIPTOMES {
     )
     ch_versions = ch_versions.mix(TRANSCRIPTOME.out.versions)
 
-    TRANSCRIPTOME.out.gffread_fasta.ifEmpty {
-        error 'No transcriptome fasta file produced.'
-    }
-
     emit:
     fasta = TRANSCRIPTOME.out.gffread_fasta
     gtf   = EXCLUDE_OVERLONG_TRANSCRIPTS.out.output

subworkflows/local/prepare_genome.nf

@@ -108,12 +108,10 @@ workflow PREPARE_GENOME {
     SAMTOOLS_FAIDX(ch_fasta, [[], []])
     ch_versions = ch_versions.mix(SAMTOOLS_FAIDX.out.versions)
 
-    ch_circexplorer2_reference = Channel.empty()
-    if (detection_tools.intersect(['circexplorer2', 'mapsplice']).size() > 0) {
-        CIRCEXPLORER2_REFERENCE(UCSC_GTFTOGENEPRED.out.genepred, [], false)
-        ch_circexplorer2_reference = CIRCEXPLORER2_REFERENCE.out.output.map { _meta, file -> file }.collect()
-        ch_versions = ch_versions.mix(CIRCEXPLORER2_REFERENCE.out.versions)
-    }
+    // Circexplorer2 reference is needed for annotation
+    CIRCEXPLORER2_REFERENCE(UCSC_GTFTOGENEPRED.out.genepred, [], false)
+    ch_circexplorer2_reference = CIRCEXPLORER2_REFERENCE.out.output.map { _meta, file -> file }.collect()
+    ch_versions = ch_versions.mix(CIRCEXPLORER2_REFERENCE.out.versions)
 
     emit:
     gtf           = ch_gtf