Switch to RData based approach in ehmm

daisybio · Nov 30, 2023 · 424f200 · 424f200
1 parent 35108ec
commit 424f200
Show file tree

Hide file tree

Showing 5 changed files with 13 additions and 18 deletions.
diff --git a/bin/split_datasets.R b/bin/split_datasets.R
@@ -1,6 +1,5 @@
 #!/usr/bin/env Rscript
 
-library(ehmm)
 library(argparser)
 library(rtracklayer)
 library(GenomicRanges)

diff --git a/modules/local/ehmm/construct_model.nf b/modules/local/ehmm/construct_model.nf
@@ -4,26 +4,22 @@ process CONSTRUCT_MODEL {
     container 'registry.hub.docker.com/bigdatainbiomedicine/inspect-ehmm'
 
     input:
-        tuple path(background_model, stageAs: 'background.model'), path(background_counts, stageAs: 'background.counts'), path(background_bed, stageAs: 'background.bed')
-        tuple path(enhancer_model, stageAs: 'enhancers.model'), path(enhancers_counts, stageAs: 'enhancers.counts'), path(enhancers_bed, stageAs: 'enhancers.bed')
-        tuple path(promoter_model, stageAs: 'promoters.model'), path(promoters_counts, stageAs: 'promoters.counts'), path(promoters_bed, stageAs: 'promoters.bed')
+        path(background_model, stageAs: 'background.RData')
+        path(enhancer_model, stageAs: 'enhancers.RData')
+        path(promoter_model, stageAs: 'promoters.RData')
 
     output:
         path('model'), emit: model_dir
-        path('model/model.txt'), emit: model
+        path('model/model.RData'), emit: model
 
     script:
         """
         mkdir -p model
 
         ehmm constructModel \\
-            --model.e ${enhancer_model} \\
-            --counts.e ${enhancers_counts} \\
-            --regions.e ${enhancers_bed} \\
-            --model.p ${promoter_model} \\
-            --counts.p ${promoters_counts} \\
-            --regions.p ${promoters_bed} \\
-            --model.bg ${background_model} \\
+            --enhancers ${enhancer_model} \\
+            --promoters ${promoter_model} \\
+            --background ${background_model} \\
             --nthreads ${task.cpus} \\
             --outdir model
         """

diff --git a/modules/local/ehmm/learn_model.nf b/modules/local/ehmm/learn_model.nf
@@ -11,7 +11,7 @@ process LEARN_MODEL {
         val(pseudocount)
 
     output:
-        tuple path("${meta.model}/model.txt"), path("${meta.model}/countmatrix.txt"), path("${meta.model}/segmentation.bed"), emit: model
+        path("${meta.model}/model.RData"), emit: model
 
     script:
         marks_string = ""

diff --git a/modules/local/ehmm/split_datasets.nf b/modules/local/ehmm/split_datasets.nf
@@ -2,7 +2,7 @@ process SPLIT_DATASETS {
     tag "split_datasets"
     label 'process_high_memory'
 
-    container 'registry.hub.docker.com/bigdatainbiomedicine/inspect-ehmm'
+    container 'registry.hub.docker.com/bigdatainbiomedicine/inspect-r'
 
     input:
         tuple val(meta), path(background)

diff --git a/subworkflows/local/ehmm.nf b/subworkflows/local/ehmm.nf
@@ -112,13 +112,13 @@ workflow EHMM {
         )
 
         ch_enhancers = APPLY_MODEL.out.enhancers
-                        .map{meta, regions -> [[id: meta["state"] + "_enhancers", 
-                                                state: meta["state"], 
+                        .map{meta, regions -> [[id: meta["id"] + "_enhancers", 
+                                                state: meta["id"], 
                                                 antibody: "enhancers"], regions]}
                         .filter{meta, regions -> regions.size() > 0}
         ch_promoters = APPLY_MODEL.out.promoters
-                        .map{meta, regions -> [[id: meta["state"] + "_promoters", 
-                                                state: meta["state"], 
+                        .map{meta, regions -> [[id: meta["id"] + "_promoters", 
+                                                state: meta["id"], 
                                                 antibody: "promoters"], regions]}
                         .filter{meta, regions -> regions.size() > 0}