From 0ec5df588b0646056519825aad4bd09563266cd9 Mon Sep 17 00:00:00 2001
From: "Kevin Bonham, PhD" <kbonham@wellesley.edu>
Date: Wed, 13 Nov 2024 11:32:56 -0500
Subject: [PATCH 1/2] mp v4 working

---
 aws-params.yaml        |  4 ++--
 metaphlan_only.nf      | 18 ++++++++++++++++++
 processes/metaphlan.nf | 11 +++--------
 3 files changed, 23 insertions(+), 10 deletions(-)
 create mode 100644 metaphlan_only.nf

diff --git a/aws-params.yaml b/aws-params.yaml
index ba0ae3e..377f1ad 100644
--- a/aws-params.yaml
+++ b/aws-params.yaml
@@ -1,8 +1,8 @@
 readsdir: "s3://vkc-nextflow/rawfastq/"
 outdir: "s3://vkc-nextflow/output/"
 human_genome: "s3://biobakery-databases/kneaddata_databases/"
-metaphlan_db: "s3://biobakery-databases/metaphlan_databases/"
+metaphlan_db: "s3://biobakery-databases/metaphlan_v4_databases/"
 humann_bowtie_db: "s3://biobakery-databases/humann_databases/chocophlan"
 humann_protein_db: "s3://biobakery-databases/humann_databases/uniref"
 humann_utility_db: "s3://biobakery-databases/humann_databases/utility_mapping"
-filepattern: "*_L00{1,2,3,4}_R{1,2}_001.fastq.gz"
\ No newline at end of file
+filepattern: "*_L00{1,2,3,4}_R{1,2}_001.fastq.gz"
diff --git a/metaphlan_only.nf b/metaphlan_only.nf
new file mode 100644
index 0000000..e4f02fc
--- /dev/null
+++ b/metaphlan_only.nf
@@ -0,0 +1,18 @@
+#!/usr/bin/env nextflow
+nextflow.enable.dsl=2
+
+include { metaphlan; metaphlan_bzip } from './processes/metaphlan.nf'
+
+workflow {
+    
+    read_pairs_ch = Channel
+        .fromFilePairs(
+            [ "$params.readsdir/$params.filepattern",
+              "$params.readsdir/*_kneaddata.fastq.gz" ],
+            size:-1)
+
+    metaphlan_db      = params.metaphlan_db
+
+    metaphlan_out = metaphlan(read_pairs_ch, metaphlan_db)
+    metaphlan_bzip = metaphlan_bzip(metaphlan_out[0], metaphlan_out[4])
+}
diff --git a/processes/metaphlan.nf b/processes/metaphlan.nf
index e081e44..ad35952 100644
--- a/processes/metaphlan.nf
+++ b/processes/metaphlan.nf
@@ -4,7 +4,6 @@ process metaphlan {
 
     input:
     tuple val(sample), path(kneads)
-    path unmatched
     path metaphlan_db
 
     output:
@@ -15,13 +14,9 @@ process metaphlan {
     path "${sample}.sam"
 
     script:
-    def forward = kneads[0]
-    def reverse = kneads[1]
-    def unf = unmatched[0]
-    def unr = unmatched[1]
-
+    
     """
-    cat $forward $reverse $unf $unr > ${sample}_grouped.fastq.gz
+    cat $kneads > ${sample}_grouped.fastq.gz
     
     metaphlan ${sample}_grouped.fastq.gz ${sample}_profile.tsv \
         --bowtie2out ${sample}_bowtie2.tsv \
@@ -49,4 +44,4 @@ process metaphlan {
     """
     bzip2 -v $sam
     """
-}
\ No newline at end of file
+}

From c277c986897853faf5139fa47ab304b0eb8ea22b Mon Sep 17 00:00:00 2001
From: "Kevin Bonham, PhD" <kbonham@wellesley.edu>
Date: Thu, 2 Jan 2025 16:12:41 -0500
Subject: [PATCH 2/2] modify main and metaphlan

---
 main.nf                | 6 ------
 processes/metaphlan.nf | 5 +++--
 2 files changed, 3 insertions(+), 8 deletions(-)

diff --git a/main.nf b/main.nf
index e99896d..aa11be5 100755
--- a/main.nf
+++ b/main.nf
@@ -14,14 +14,8 @@ workflow {
 
     human_genome      = params.human_genome
     metaphlan_db      = params.metaphlan_db
-    humann_bowtie_db  = params.humann_bowtie_db
-    humann_protein_db = params.humann_protein_db
-    humann_utility_db = params.humann_utility_db
     
     knead_out     = kneaddata(read_pairs_ch, human_genome)
     metaphlan_out = metaphlan(knead_out[0], knead_out[1], metaphlan_db)
     metaphlan_bzip = metaphlan_bzip(metaphlan_out[0], metaphlan_out[4])
-    humann_out    = humann(metaphlan_out[0], metaphlan_out[1], metaphlan_out[2], humann_bowtie_db, humann_protein_db)
-    regroup_out   = humann_regroup(humann_out[0], humann_out[1], humann_utility_db)
-    humann_rename(regroup_out, humann_utility_db)
 }
diff --git a/processes/metaphlan.nf b/processes/metaphlan.nf
index ad35952..ecc1a40 100644
--- a/processes/metaphlan.nf
+++ b/processes/metaphlan.nf
@@ -3,8 +3,9 @@ process metaphlan {
     publishDir "$params.outdir/metaphlan", pattern: "{*.tsv}"
 
     input:
-    tuple val(sample), path(kneads)
-    path metaphlan_db
+        tuple val(sample), path(kneads)
+        path unmatched
+        path metaphlan_db
 
     output:
     val  sample                  , emit: sample