From f1d2b5acaeabf6aff3f9be494dc42f8b4591cc96 Mon Sep 17 00:00:00 2001
From: Edmund Miller <git@edmundmiller.dev>
Date: Sat, 16 Nov 2024 15:12:56 -0600
Subject: [PATCH] refactor: Move uniqmap generation to references

https://github.com/nf-core/references/issues/3
---
 main.nf                                   |  1 +
 modules/nf-core/homer/findpeaks/main.nf   |  3 +-
 nextflow_schema.json                      |  6 ++++
 subworkflows/nf-core/homer/groseq/main.nf | 42 ++++++-----------------
 workflows/nascent.nf                      |  2 ++
 5 files changed, 22 insertions(+), 32 deletions(-)

diff --git a/main.nf b/main.nf
index 4e0cbc32..e698542d 100644
--- a/main.nf
+++ b/main.nf
@@ -69,6 +69,7 @@ workflow NFCORE_NASCENT {
         params.bowtie2_index,
         params.hisat2_index,
         params.star_index,
+        params.homer_uniqmap
     )
     emit:
     multiqc_report = NASCENT.out.multiqc_report // channel: /path/to/multiqc_report.html
diff --git a/modules/nf-core/homer/findpeaks/main.nf b/modules/nf-core/homer/findpeaks/main.nf
index 71e26679..bce67462 100644
--- a/modules/nf-core/homer/findpeaks/main.nf
+++ b/modules/nf-core/homer/findpeaks/main.nf
@@ -23,13 +23,14 @@ process HOMER_FINDPEAKS {
     def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
     def VERSION = '4.11' // WARN: Version information not provided by tool on CLI. Please update this string when bumping container versions.
+    def uniqmap_flag = uniqmap ? "-uniqmap $uniqmap" : ""
     """
 
     findPeaks \\
         $tagDir \\
         $args \\
         -o ${prefix}.peaks.txt \\
-        -uniqmap $uniqmap
+        $uniqmap_flag
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
diff --git a/nextflow_schema.json b/nextflow_schema.json
index 60465a52..ba327409 100644
--- a/nextflow_schema.json
+++ b/nextflow_schema.json
@@ -254,6 +254,12 @@
                     "fa_icon": "fas fa-copy",
                     "help_text": "> **NB** If none provided, index will be generated automatically from the FASTA reference."
                 },
+                "homer_uniqmap": {
+                    "type": "string",
+                    "description": "Path to HOMER uniqmap file or URL to download.",
+                    "fa_icon": "fas fa-file-archive",
+                    "help_text": "> **NB** If none provided, will be downloaded automatically from the HOMER website. See nf-core/references for generation"
+                },
                 "save_reference": {
                     "type": "boolean",
                     "description": "If generated by the pipeline save the BWA index in the results directory.",
diff --git a/subworkflows/nf-core/homer/groseq/main.nf b/subworkflows/nf-core/homer/groseq/main.nf
index 4df5b089..68afb1fc 100644
--- a/subworkflows/nf-core/homer/groseq/main.nf
+++ b/subworkflows/nf-core/homer/groseq/main.nf
@@ -2,49 +2,29 @@
  * Identify transcripts with homer
  */
 
-include { HOMER_GETMAPPABLEREGIONS } from '../../../../modules/local/homer/getmappableregions/main'
-include { HOMER_CREATEUNIQMAP      } from '../../../../modules/local/homer/createuniqmap/main'
-include { UNZIP                    } from '../../../../modules/nf-core/unzip/main'
+include { UNZIP                  } from '../../../../modules/nf-core/unzip/main'
 
-include { HOMER_MAKETAGDIRECTORY   } from '../../../../modules/nf-core/homer/maketagdirectory/main'
-include { HOMER_MAKEUCSCFILE       } from '../../../../modules/nf-core/homer/makeucscfile/main'
-include { HOMER_FINDPEAKS          } from '../../../../modules/nf-core/homer/findpeaks/main'
-include { HOMER_POS2BED            } from '../../../../modules/nf-core/homer/pos2bed/main'
+include { HOMER_MAKETAGDIRECTORY } from '../../../../modules/nf-core/homer/maketagdirectory/main'
+include { HOMER_MAKEUCSCFILE     } from '../../../../modules/nf-core/homer/makeucscfile/main'
+include { HOMER_FINDPEAKS        } from '../../../../modules/nf-core/homer/findpeaks/main'
+include { HOMER_POS2BED          } from '../../../../modules/nf-core/homer/pos2bed/main'
 
 workflow HOMER_GROSEQ {
     take:
-    bam   // channel: [ val(meta), [ reads ] ]
-    fasta //    file: /path/to/bwa/index/
+    bam     // channel: [ val(meta), [ reads ] ]
+    fasta   //    file: /path/to/bwa/index/
+    uniqmap
 
     main:
 
     ch_versions = Channel.empty()
 
     ch_uniqmap = Channel.empty()
-    if (!params.homer_uniqmap) {
-        // Split FASTA by chromosome
-        split_fastas = fasta
-            .splitFasta(by: 1, file: true)
-            .toSortedList()
-
-        // Generate mappable regions
-        HOMER_GETMAPPABLEREGIONS(
-            split_fastas,
-            1000000000,
-            50
-        )
-        ch_versions = ch_versions.mix(HOMER_GETMAPPABLEREGIONS.out.versions)
-
-        // Create uniqmap directory
-        HOMER_CREATEUNIQMAP(
-            HOMER_GETMAPPABLEREGIONS.out.txt
-        )
-        ch_versions = ch_versions.mix(HOMER_CREATEUNIQMAP.out.versions)
-
-        ch_uniqmap = HOMER_CREATEUNIQMAP.out.uniqmap_dir
+    if (uniqmap.endsWith('.zip')) {
+        ch_uniqmap = UNZIP([[:], params.homer_uniqmap]).map { it[1] }
     }
     else {
-        ch_uniqmap = UNZIP([[:], params.homer_uniqmap]).map { it[1] }
+        ch_uniqmap = uniqmap
     }
 
     /*
diff --git a/workflows/nascent.nf b/workflows/nascent.nf
index 0eb1b33e..af9b1996 100644
--- a/workflows/nascent.nf
+++ b/workflows/nascent.nf
@@ -59,6 +59,7 @@ workflow NASCENT {
     ch_bowtie2_index
     ch_hisat2_index
     ch_star_index
+    ch_uniqmap
 
 
     main:
@@ -294,6 +295,7 @@ workflow NASCENT {
         ch_gxf,
         PREPARE_GENOME.out.fasta,
         PREPARE_GENOME.out.chrom_sizes,
+        ch_uniqmap
     )
     ch_grohmm_multiqc = TRANSCRIPT_INDENTIFICATION.out.grohmm_td_plot.collect()
     ch_homer_multiqc = TRANSCRIPT_INDENTIFICATION.out.homer_peaks