Merge pull request nf-core#78 from LouisLeNezet/validate

Validate
LouisLeNezet · Jun 7, 2024 · ec11c67 · ec11c67
2 parents cf7a7fc + aa46859
commit ec11c67
Show file tree

Hide file tree

Showing 7 changed files with 11 additions and 11 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -29,6 +29,7 @@ Initial release of nf-core/phaseimpute, created with the [nf-core](https://nf-co
 - [#51](https://github.com/nf-core/phaseimpute/pull/51) - Update all process and fix linting errors. Remove fastqc added by the template.
 - [#56](https://github.com/nf-core/phaseimpute/pull/56) - Move to nf-test to check the output files names generated. Fix validation and concatenation by chromosomes missing. Add dedicated GLIMPSE1 subworkflow. Fix posfile generation to be done once for glimpse and stitch.
 - [#68](https://github.com/nf-core/phaseimpute/pull/68) - QUILT can handle external params chunks and hap-legend files.
+- [#78](https://github.com/nf-core/phaseimpute/pull/78) - Separate validate step from panel preparation.
 
 ### `Fixed`
 

diff --git a/assets/schema_posfile.json b/assets/schema_posfile.json
@@ -11,7 +11,7 @@
                 "type": "string",
                 "pattern": "^\\S+$",
                 "errorMessage": "Panel name must be provided as a string and cannot contain spaces",
-                "meta": ["panel"]
+                "meta": ["id"]
             },
             "chr": {
                 "type": "string",

diff --git a/conf/test_validate.config b/conf/test_validate.config
@@ -25,10 +25,9 @@ params {
     input_region         = "${projectDir}/tests/csv/region.csv"
 
     // Genome references
-    fasta  = params.pipelines_testdata_base_path + "reference_genome/21_22/hs38DH.chr21_22.fa"
-    panel  = "${projectDir}/tests/csv/panel.csv"
-    phased = true
-    map    = "${projectDir}/tests/csv/map.csv"
+    fasta   = params.pipelines_testdata_base_path + "reference_genome/21_22/hs38DH.chr21_22.fa"
+    posfile = "${projectDir}/tests/csv/posfile.csv"
+    map     = "${projectDir}/tests/csv/map.csv"
 
     // Pipeline steps
     steps   = "validate"

diff --git a/docs/usage.md b/docs/usage.md
@@ -320,12 +320,13 @@ Make sure the csv with the input panel is the output from `--step panelprep` or
 ### Start with validation `--steps validate`
 
 This steps compares a _truth_ VCF to an _imputed_ VCF in order to compute imputation accuracy.
+This also needs the frequency of the alleles. They can be computed from the reference panel by running the `--steps panelprep` and using the `--panel` with the `--compute_freq` flag ; or by using `--posfile samplesheet.csv`.
 
 ```bash
 nextflow run nf-core/phaseimpute --input samplesheet.csv --input_truth truth.csv --steps validate --outdir results --genome GRCh37 -profile docker
 ```
 
-The required flags for this mode are:
+The required flags for this mode only are:
 
 - `--steps validate`: The steps to run.
 - `--input samplesheet.csv`: The samplesheet containing the input sample files in `vcf` format.

diff --git a/subworkflows/local/utils_nfcore_phaseimpute_pipeline/main.nf b/subworkflows/local/utils_nfcore_phaseimpute_pipeline/main.nf
@@ -216,8 +216,7 @@ workflow PIPELINE_INITIALISATION {
     //
     if (params.posfile) {
         ch_posfile = Channel
-                    .fromSamplesheet("posfile")
-                    .map {meta, vcf, csi, txt -> [ meta, vcf, csi, txt ]}
+            .fromSamplesheet("posfile")
     } else {
         ch_posfile = [[],[]]
     }

diff --git a/tests/csv/posfile.csv b/tests/csv/posfile.csv
@@ -1,3 +1,3 @@
 panel,chr,vcf,index,txt
 1000GP.s.norel,chr21,"https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/21/1000GP.chr21.s.norel.sites.vcf.gz","https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/21/1000GP.chr21.s.norel.sites.vcf.gz.csi","https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/21/1000GP.chr21.s.norel.tsv.gz"
-1000GP.s.norel,chr22,"https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/22/1000GP.chr22.s.norel.sites.vcf.gz","https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/21/1000GP.chr21.s.norel.sites.vcf.gz.csi","https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/22/1000GP.chr22.s.norel.tsv.gz"
+1000GP.s.norel,chr22,"https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/22/1000GP.chr22.s.norel.sites.vcf.gz","https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/22/1000GP.chr22.s.norel.sites.vcf.gz.csi","https://raw.githubusercontent.com/nf-core/test-datasets/phaseimpute/data/panel/22/1000GP.chr22.s.norel.tsv.gz"
diff --git a/workflows/phaseimpute/main.nf b/workflows/phaseimpute/main.nf
@@ -83,7 +83,7 @@ workflow PHASEIMPUTE {
     ch_region               // channel: region to use [ [chr, region], region]
     ch_depth                // channel: depth select  [ [depth], depth ]
     ch_map                  // channel: genetic map   [ [chr], map]
-    ch_posfile              // channel: posfile       [ [chr], txt]
+    ch_posfile              // channel: posfile       [ [chr], vcf, txt]
     ch_chunks               // channel: chunks        [ [chr], txt]
     ch_versions             // channel: versions of software used
 
@@ -135,7 +135,7 @@ workflow PHASEIMPUTE {
     //
     // Prepare panel
     //
-    if (params.steps.split(',').contains("panelprep") || params.steps.split(',').contains("validate") || params.steps.split(',').contains("all")) {
+    if (params.steps.split(',').contains("panelprep") || params.steps.split(',').contains("all")) {
         // Check chr prefix and remove if necessary
         VCF_CHR_CHECK(ch_panel, ch_fasta)
         ch_versions = ch_versions.mix(VCF_CHR_CHECK.out.versions)