feat: one dna or RNA count file across multiple replicates (#144)

* feat: using one experiment across all replicates for DNA (or RNA)

* snakemake format

---------

Co-authored-by: Max Schubach <max.schubach@bih-charite.de>

workflow/rules/assigned_counts.smk

@@ -62,7 +62,9 @@ rule assigned_counts_assignBarcodes:
     conda:
         "../envs/python3.yaml"
     input:
-        counts=lambda wc: getFinalCounts(wc.project, wc.config, wc.type, "counts"),
+        counts=lambda wc: getFinalCounts(
+            wc.project, wc.config, wc.condition, wc.type, "counts"
+        ),
         association="results/experiments/{project}/assignment/{assignment}.tsv.gz",
         script=getScript("count/merge_BC_and_assignment.py"),
     output:
@@ -213,7 +215,8 @@ rule assigned_counts_combine_replicates_barcode_output:
         ]["bc_threshold"],
         bc_counts=lambda wc: " ".join(
             [
-                "--counts %s results/experiments/%s/assigned_counts/%s/%s/%s_%s_barcode_assigned_counts.tsv.gz" % (rep, wc.project, wc.assignment, wc.config, wc.condition, rep)
+                "--counts %s results/experiments/%s/assigned_counts/%s/%s/%s_%s_barcode_assigned_counts.tsv.gz"
+                % (rep, wc.project, wc.assignment, wc.config, wc.condition, rep)
                 for rep in getReplicatesOfCondition(wc.project, wc.condition)
             ]
         ),

workflow/rules/common.smk

@@ -276,15 +276,16 @@ def getOutputConditionReplicateType_helper(files, project, skip={}):
             return []
     conditions = getConditions(project)
     for condition in conditions:
-        replicates = getReplicatesOfCondition(project, condition)
         for file in files:
-            output += expand(
-                file,
-                project=project,
-                condition=condition,
-                replicate=replicates,
-                type=["RNA", "DNA"],
-            )
+            for type in ["DNA", "RNA"]:
+                replicates = getReplicatesOfConditionType(project, condition, type)
+                output += expand(
+                    file,
+                    project=project,
+                    condition=condition,
+                    replicate=replicates,
+                    type=type,
+                )
     return output
 
 
@@ -688,24 +689,46 @@ def counts_getSamplingConfig(project, conf, dna_or_rna, command):
     return ""
 
 
-def getFinalCounts(project, conf, rna_or_dna, raw_or_assigned):
+def getReplicatesOfConditionType(project, condition, rna_or_dna):
+    exp = getExperiments(project)
+
+    replicates = getReplicatesOfCondition(project, condition)
+
+    if f"{rna_or_dna}_BC_F" in exp.columns:
+
+        exp_filter = exp[exp.Condition == condition]
+
+        if len(replicates) > 1 and exp_filter[f"{rna_or_dna}_BC_F"].nunique() == 1:
+            return [replicates[0]]
+
+    return replicates
+
+
+def getFinalCounts(project, conf, condition, rna_or_dna, raw_or_assigned):
     output = ""
+
+    replicates = getReplicatesOfConditionType(project, condition, rna_or_dna)
+    if len(replicates) > 1:
+        replicate = "{replicate}"
+    else:
+        replicate = replicates[0]
+
     if raw_or_assigned == "counts":
         if useSampling(project, conf, rna_or_dna):
             output = (
-                "results/experiments/{project}/%s/{condition}_{replicate}_%s_final_counts.sampling.{config}.tsv.gz"
-                % (raw_or_assigned, rna_or_dna)
+                "results/experiments/{project}/%s/{condition}_%s_%s_final_counts.sampling.{config}.tsv.gz"
+                % (raw_or_assigned, replicate, rna_or_dna)
             )
 
         else:
             output = (
-                "results/experiments/{project}/%s/{condition}_{replicate}_%s_final_counts.tsv.gz"
-                % (raw_or_assigned, rna_or_dna)
+                "results/experiments/{project}/%s/{condition}_%s_%s_final_counts.tsv.gz"
+                % (raw_or_assigned, replicate, rna_or_dna)
             )
     else:
         output = (
-            "results/experiments/{project}/%s/{condition}_{replicate}_%s_final_counts.config.{config}.tsv.gz"
-            % (raw_or_assigned, rna_or_dna)
+            "results/experiments/{project}/%s/{condition}_%s_%s_final_counts.config.{config}.tsv.gz"
+            % (raw_or_assigned, replicate, rna_or_dna)
         )
     return output
 
@@ -733,20 +756,6 @@ def assignedCounts_getAssignmentSamplingConfig(project, assignment, command):
 # statistic.smk specific functions
 
 
-# get all counts of experiment (rule statistic_counts)
-def getCountStats(project, countType):
-    exp = getExperiments(project)
-    output = []
-    for index, row in exp.iterrows():
-        output += expand(
-            "results/experiments/{{project}}/statistic/counts/{condition}_{replicate}_{type}_{{countType}}_counts.tsv.gz",
-            condition=row["Condition"],
-            replicate=row["Replicate"],
-            type=["DNA", "RNA"],
-        )
-    return output
-
-
 # get all barcodes of experiment (rule statistic_BC_in_RNA_DNA)
 def getBCinRNADNAStats(wc):
     exp = getExperiments(wc.project)

workflow/rules/counts.smk

@@ -110,8 +110,12 @@ rule counts_dna_rna_merge_counts:
     conda:
         "../envs/default.yaml"
     input:
-        dna=lambda wc: getFinalCounts(wc.project, wc.config, "DNA", wc.raw_or_assigned),
-        rna=lambda wc: getFinalCounts(wc.project, wc.config, "RNA", wc.raw_or_assigned),
+        dna=lambda wc: getFinalCounts(
+            wc.project, wc.config, wc.condition, "DNA", wc.raw_or_assigned
+        ),
+        rna=lambda wc: getFinalCounts(
+            wc.project, wc.config, wc.condition, "RNA", wc.raw_or_assigned
+        ),
     output:
         "results/experiments/{project}/{raw_or_assigned}/{condition}_{replicate}.merged.config.{config}.tsv.gz",
     params:

workflow/rules/statistic/bc_overlap.smk

@@ -8,7 +8,9 @@ rule statistic_bc_overlap_run:
         "../../envs/r.yaml"
     input:
         files=lambda wc: expand(
-            getFinalCounts(wc.project, wc.config, wc.type, wc.raw_or_assigned),
+            getFinalCounts(
+                wc.project, wc.config, wc.condition, wc.type, wc.raw_or_assigned
+            ),
             project=wc.project,
             condition=wc.condition,
             config=wc.config,
@@ -22,7 +24,9 @@ rule statistic_bc_overlap_run:
     params:
         input=lambda wc: ",".join(
             expand(
-                getFinalCounts(wc.project, wc.config, wc.type, wc.raw_or_assigned),
+                getFinalCounts(
+                    wc.project, wc.config, wc.condition, wc.type, wc.raw_or_assigned
+                ),
                 project=wc.project,
                 condition=wc.condition,
                 config=wc.config,
@@ -31,7 +35,7 @@ rule statistic_bc_overlap_run:
         ),
         cond="{condition}_{type}",
         replicates=lambda wc: ",".join(
-            getReplicatesOfCondition(wc.project, wc.condition)
+            getReplicatesOfConditionType(wc.project, wc.condition, wc.type)
         ),
     log:
         temp(

workflow/rules/statistic/counts.smk

@@ -1,3 +1,6 @@
+include: "counts_common.smk"
+
+
 #################################
 ## count 10 most frequent UMIs ##
 #################################
@@ -148,15 +151,11 @@ rule statistic_counts_BC_in_RNA_DNA:
     conda:
         "../../envs/default.yaml"
     input:
-        dna=lambda wc: (
-            "results/experiments/{project}/counts/{condition}_{replicate}_DNA_{countType}_counts.tsv.gz"
-            if wc.countType != "raw"
-            else getRawCounts(wc.project, "DNA")
+        dna=lambda wc: statistic_counts_BC_in_RNA_DNA_helper(
+            project, wc.condition, "DNA", wc.countType
         ),
-        rna=lambda wc: (
-            "results/experiments/{project}/counts/{condition}_{replicate}_RNA_{countType}_counts.tsv.gz"
-            if wc.countType != "raw"
-            else getRawCounts(wc.project, "RNA")
+        rna=lambda wc: statistic_counts_BC_in_RNA_DNA_helper(
+            project, wc.condition, "RNA", wc.countType
         ),
     output:
         temp(

workflow/rules/statistic/counts_common.smk

@@ -0,0 +1,37 @@
+def statistic_counts_BC_in_RNA_DNA_helper(project, condition, dna_or_rna, countType):
+
+    replicates = getReplicatesOfConditionType(project, condition, dna_or_rna)
+
+    if countType == "raw":
+        output = getRawCounts(project, dna_or_rna)
+    else:
+        output = (
+            "results/experiments/{project}/counts/{condition}_{replicate}_%s_{countType}_counts.tsv.gz"
+            % dna_or_rna
+        )
+
+    if len(replicates) == 1:
+        output = output.replace("{replicate}", replicates[0])
+
+    return output
+
+
+# get all counts of experiment (rule statistic_counts)
+def getCountStats(project, countType):
+    exp = getExperiments(project)
+    output = []
+    for index, row in exp.iterrows():
+        condition = row["Condition"]
+        for dna_or_rna in ["DNA", "RNA"]:
+            replicates = getReplicatesOfConditionType(project, condition, dna_or_rna)
+            if len(replicates) == 1:
+                replicate = replicates[0]
+            else:
+                replicate = row["Replicate"]
+            output += expand(
+                "results/experiments/{{project}}/statistic/counts/{condition}_{replicate}_{type}_{{countType}}_counts.tsv.gz",
+                condition=condition,
+                replicate=replicate,
+                type=dna_or_rna,
+            )
+    return output