Merge pull request #18 from wslh-bio/dev/4.0

Dev/4.0
wslh-bio · Aug 28, 2024 · 3bb7e98 · 3bb7e98
2 parents f6cce71 + 1b294e5
commit 3bb7e98
Show file tree

Hide file tree

Showing 6 changed files with 120 additions and 11 deletions.
diff --git a/README.md b/README.md
@@ -46,6 +46,19 @@ nextflow run wslh-bio/dryad \
    --fasta <REFERENCE_FASTA> \
    --alignment_based 
 ```
+
+To run both and alignment based and an alignment free comparison, use:
+```bash
+nextflow run wslh-bio/dryad \
+   -latest \
+   -profile <docker/singularity/.../institute> \
+   --input samplesheet.csv \
+   --outdir <OUTDIR> \
+   --fasta <REFERENCE_FASTA> \
+   --alignment_based \
+   --alignment_free
+```
+
 * Nextflow caches previously run pipelines. This can result in an older version of a pipeline being utilized. To get the most up-to-date version of a pipeline like Dryad, use the `-latest` tag.
 
 ## Input

diff --git a/conf/test.config b/conf/test.config
@@ -19,10 +19,12 @@ params {
     max_memory = '6.GB'
     max_time   = '6.h'
 
-    // Input data
-    // TODO nf-core: Specify the paths to your test data on nf-core/test-datasets
-    // TODO nf-core: Give any required params for the test so that command line flags are not needed
-    input  = 'https://raw.githubusercontent.com/nf-core/test-datasets/viralrecon/samplesheet/samplesheet_test_illumina_amplicon.csv'
+    // Input data for full size test
+    input = 'https://raw.githubusercontent.com/wslh-bio/dryad/main/samplesheets/test.csv'
 
-
+    // Fasta references
+    fasta = 'https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/reference/England1.fna'
+
+    // Alignment
+    alignment_based = true
 }
diff --git a/conf/test_full.config b/conf/test_full.config
@@ -15,10 +15,12 @@ params {
     config_profile_description = 'Full test dataset to check pipeline function'
 
     // Input data for full size test
-    // TODO nf-core: Specify the paths to your full test data ( on nf-core/test-datasets or directly in repositories, e.g. SRA)
-    // TODO nf-core: Give any required params for the test so that command line flags are not needed
-    input = 'https://raw.githubusercontent.com/nf-core/test-datasets/viralrecon/samplesheet/samplesheet_full_illumina_amplicon.csv'
+    input = 'https://raw.githubusercontent.com/wslh-bio/dryad/main/samplesheets/test_full.csv'
 
     // Fasta references
-    fasta = 'https://raw.githubusercontent.com/nf-core/test-datasets/viralrecon/genome/NC_045512.2/GCF_009858895.2_ASM985889v3_genomic.200409.fna.gz'
+    fasta = 'https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/reference/England1.fna'
+
+    // Alignment
+    alignment_based = true
+    alignment_free = true
 }
diff --git a/samplesheets/test.csv b/samplesheets/test.csv
@@ -0,0 +1,46 @@
+sample,fasta
+Al-Hasa_12_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_12_2013.fna
+Al-Hasa_15_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_15_2013.fna
+Al-Hasa_16_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_16_2013.fna
+Al-Hasa_17_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_17_2013.fna
+Al-Hasa_18_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_18_2013.fna
+Al-Hasa_19_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_19_2013.fna
+Al-Hasa_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_1_2013.fna
+Al-Hasa_21_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_21_2013.fna
+Al-Hasa_25_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_25_2013.fna
+Al-Hasa_2_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_2_2013.fna
+Al-Hasa_3_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_3_2013.fna
+Al-Hasa_4_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_4_2013.fna
+Bisha_1_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Bisha_1_2012.fna
+Buraidah_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Buraidah_1_2013.fna
+EMC_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/EMC_2012.fna
+England-Qatar_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/England-Qatar_2012.fna
+FRA-UAE,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/FRA-UAE.fna
+Hafr-Al-Batin_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Hafr-Al-Batin_1_2013.fna
+Hafr-Al-Batin_2_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Hafr-Al-Batin_2_2013.fna
+Hafr-Al_Batin_6_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Hafr-Al_Batin_6_2013.fna
+Indiana-USA-1_Saudi_Arabia_2014,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Indiana-USA-1_Saudi_Arabia_2014.fna
+Jeddah_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Jeddah_1_2013.fna
+Jordan-N3_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Jordan-N3_2012.fna
+KF192507.1,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KF192507.1.fna
+KFU-HKU_1,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KFU-HKU_1.fna
+KFU-HKU_13,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KFU-HKU_13.fna
+KFU-HKU_19Dam,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KFU-HKU_19Dam.fna
+KJ477102.1,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KJ477102.1.fna
+KSA-CAMEL-363,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-363.fna
+KSA-CAMEL-376,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-376.fna
+KSA-CAMEL-378,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-378.fna
+KSA-CAMEL-503,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-503.fna
+KSA-CAMEL-505,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-505.fna
+NC_019843.2,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/NC_019843.2.fna
+Qatar3,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Qatar3.fna
+Qatar4,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Qatar4.fna
+Riyadh_14_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_14_2013.fna
+Riyadh_1_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_1_2012.fna
+Riyadh_2_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_2_2012.fna
+Riyadh_3_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_3_2013.fna
+Riyadh_4_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_4_2013.fna
+Riyadh_5_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_5_2013.fna
+Riyadh_9_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_9_2013.fna
+Taif_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Taif_1_2013.fna
+Wadi-Ad-Dawasir_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Wadi-Ad-Dawasir_1_2013.fna
diff --git a/samplesheets/test_full.csv b/samplesheets/test_full.csv
@@ -0,0 +1,46 @@
+sample,fasta
+Al-Hasa_12_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_12_2013.fna
+Al-Hasa_15_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_15_2013.fna
+Al-Hasa_16_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_16_2013.fna
+Al-Hasa_17_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_17_2013.fna
+Al-Hasa_18_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_18_2013.fna
+Al-Hasa_19_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_19_2013.fna
+Al-Hasa_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_1_2013.fna
+Al-Hasa_21_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_21_2013.fna
+Al-Hasa_25_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_25_2013.fna
+Al-Hasa_2_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_2_2013.fna
+Al-Hasa_3_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_3_2013.fna
+Al-Hasa_4_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Al-Hasa_4_2013.fna
+Bisha_1_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Bisha_1_2012.fna
+Buraidah_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Buraidah_1_2013.fna
+EMC_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/EMC_2012.fna
+England-Qatar_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/England-Qatar_2012.fna
+FRA-UAE,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/FRA-UAE.fna
+Hafr-Al-Batin_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Hafr-Al-Batin_1_2013.fna
+Hafr-Al-Batin_2_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Hafr-Al-Batin_2_2013.fna
+Hafr-Al_Batin_6_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Hafr-Al_Batin_6_2013.fna
+Indiana-USA-1_Saudi_Arabia_2014,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Indiana-USA-1_Saudi_Arabia_2014.fna
+Jeddah_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Jeddah_1_2013.fna
+Jordan-N3_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Jordan-N3_2012.fna
+KF192507.1,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KF192507.1.fna
+KFU-HKU_1,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KFU-HKU_1.fna
+KFU-HKU_13,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KFU-HKU_13.fna
+KFU-HKU_19Dam,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KFU-HKU_19Dam.fna
+KJ477102.1,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KJ477102.1.fna
+KSA-CAMEL-363,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-363.fna
+KSA-CAMEL-376,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-376.fna
+KSA-CAMEL-378,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-378.fna
+KSA-CAMEL-503,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-503.fna
+KSA-CAMEL-505,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/KSA-CAMEL-505.fna
+NC_019843.2,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/NC_019843.2.fna
+Qatar3,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Qatar3.fna
+Qatar4,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Qatar4.fna
+Riyadh_14_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_14_2013.fna
+Riyadh_1_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_1_2012.fna
+Riyadh_2_2012,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_2_2012.fna
+Riyadh_3_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_3_2013.fna
+Riyadh_4_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_4_2013.fna
+Riyadh_5_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_5_2013.fna
+Riyadh_9_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Riyadh_9_2013.fna
+Taif_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Taif_1_2013.fna
+Wadi-Ad-Dawasir_1_2013,https://raw.githubusercontent.com/wslh-bio/dryad/main/test-dataset/short_dataset/genomes/Wadi-Ad-Dawasir_1_2013.fna
diff --git a/workflows/dryad.nf b/workflows/dryad.nf
@@ -40,7 +40,7 @@ workflow DRYAD {
     //
     // Error Handling
     //
-    if (params.alignment_free && params.fasta) {
+    if (params.alignment_free && params.fasta && !params.alignment_based) {
         error("ERROR: An alignment free comparison does not use a reference fasta. Do you want to run an alignment based comparison instead?\nDryad terminating...")
         exit(1)
     }
@@ -94,7 +94,7 @@ workflow DRYAD {
     //
     // SUBWORKFLOW: Alignment Free
     //
-    if (params.alignment_free && !params.fasta) {
+    if (params.alignment_free) {
         ALIGNMENT_FREE (
             ch_for_alignments,
             params.task.cpus