Allow transcriptome-only salmon indexing

modules/nf-core/salmon/index/main.nf

@@ -20,22 +20,29 @@ process SALMON_INDEX {
 
     script:
     def args = task.ext.args ?: ''
-    def get_decoy_ids = "grep '^>' $genome_fasta | cut -d ' ' -f 1 | cut -d \$'\\t' -f 1 > decoys.txt"
-    def gentrome      = "gentrome.fa"
-    if (genome_fasta.endsWith('.gz')) {
-        get_decoy_ids = "grep '^>' <(gunzip -c $genome_fasta) | cut -d ' ' -f 1 | cut -d \$'\\t' -f 1 > decoys.txt"
-        gentrome      = "gentrome.fa.gz"
+    def decoys = ''
+    def fasta = transcript_fasta
+    if (genome_fasta){
+        if (genome_fasta.endsWith('.gz')) {
+            genome_fasta = "<(gunzip -c $genome_fasta)"
+        }
+        decoys='-d decoys.txt'
+        fasta='gentrome.fa'
+    }
+    if (transcript_fasta.endsWith('.gz')) {
+        transcript_fasta = "<(gunzip -c $transcript_fasta)"
     }
     """
-    $get_decoy_ids
-    sed -i.bak -e 's/>//g' decoys.txt
-    cat $transcript_fasta $genome_fasta > $gentrome
+    if [ -n '$genome_fasta' ]; then
+        grep '^>' $genome_fasta | cut -d ' ' -f 1 | cut -d \$'\\t' -f 1 | sed 's/>//g' > decoys.txt
+        cat $transcript_fasta $genome_fasta > $fasta
+    fi
 
     salmon \\
         index \\
         --threads $task.cpus \\
-        -t $gentrome \\
-        -d decoys.txt \\
+        -t $fasta \\
+        $decoys \\
         $args \\
         -i salmon
 

modules/nf-core/salmon/index/tests/main.nf.test

@@ -26,13 +26,43 @@ nextflow_process {
             assertAll(
                 { assert process.success },
                 { assert path(process.out.index.get(0)).exists() },
-                { assert snapshot(process.out.versions).match("versions") }
+                { assert snapshot(
+                    file(process.out.index[0]).listFiles().collect { it.getName() }.sort().toString(),
+                    process.out.versions
+                ).match()}
             )
         }
 
     }
 
-     test("sarscov2 stub") {
+    test("sarscov2 transcriptome only") {
+
+        when {
+            params {
+                outdir = "$outputDir"
+            }
+            process {
+                """
+                input[0] = Channel.of([])
+                input[1] = Channel.of([file(params.modules_testdata_base_path + "genomics/sarscov2/genome/transcriptome.fasta", checkIfExists: true)])
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert path(process.out.index.get(0)).exists() },
+                { assert snapshot(
+                    file(process.out.index[0]).listFiles().collect { it.getName() }.sort().toString(),
+                    process.out.versions
+                ).match()}
+            )
+        }
+
+    }
+
+    test("sarscov2 stub") {
         options "-stub"
         when {
             params {
@@ -50,7 +80,10 @@ nextflow_process {
             assertAll(
                 { assert process.success },
                 { assert path(process.out.index.get(0)).exists() },
-                { assert snapshot(process.out.versions).match("versions stub") }
+                { assert snapshot(
+                    file(process.out.index[0]).listFiles().collect { it.getName() }.sort().toString(),
+                    process.out.versions
+                ).match()}
             )
         }
 

modules/nf-core/salmon/index/tests/main.nf.test.snap

@@ -1,26 +1,41 @@
 {
-    "versions": {
+    "sarscov2 stub": {
         "content": [
+            "[complete_ref_lens.bin, ctable.bin, ctg_offsets.bin, duplicate_clusters.tsv, info.json, mphf.bin, pos.bin, pre_indexing.log, rank.bin, refAccumLengths.bin, ref_indexing.log, reflengths.bin, refseq.bin, seq.bin, versionInfo.json]",
             [
                 "versions.yml:md5,85337fa0a286ea35073ee5260974e307"
             ]
         ],
         "meta": {
-            "nf-test": "0.9.0",
-            "nextflow": "24.04.4"
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.2"
         },
-        "timestamp": "2024-10-18T10:00:47.087293189"
+        "timestamp": "2025-01-20T12:57:51.498323"
     },
-    "versions stub": {
+    "sarscov2": {
         "content": [
+            "[complete_ref_lens.bin, ctable.bin, ctg_offsets.bin, duplicate_clusters.tsv, info.json, mphf.bin, pos.bin, pre_indexing.log, rank.bin, refAccumLengths.bin, ref_indexing.log, reflengths.bin, refseq.bin, seq.bin, versionInfo.json]",
             [
                 "versions.yml:md5,85337fa0a286ea35073ee5260974e307"
             ]
         ],
         "meta": {
-            "nf-test": "0.9.0",
-            "nextflow": "24.04.4"
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.2"
         },
-        "timestamp": "2024-10-18T10:01:03.89824494"
+        "timestamp": "2025-01-20T12:57:33.474302"
+    },
+    "sarscov2 transcriptome only": {
+        "content": [
+            "[complete_ref_lens.bin, ctable.bin, ctg_offsets.bin, duplicate_clusters.tsv, info.json, mphf.bin, pos.bin, pre_indexing.log, rank.bin, refAccumLengths.bin, ref_indexing.log, reflengths.bin, refseq.bin, seq.bin, versionInfo.json]",
+            [
+                "versions.yml:md5,85337fa0a286ea35073ee5260974e307"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.2"
+        },
+        "timestamp": "2025-01-20T12:57:42.420247"
     }
 }