fix: Move signed_pi_value columns to the end (again) (#117)

- Undo my overwriting of the position of the signed_pi_cols
- Small bug fixes

---------

Co-authored-by: Addimator <adpri100hhu.de>

workflow/resources/datavzrd/diffexp-template.yaml

@@ -134,7 +134,7 @@ views:
  display-mode: hidden
  regex('qval_.+'):
  display-mode: hidden
- regex('signed_pi_value_.+'):
+ ?f"regex('signed_pi_value_{params.primary_variable}.+')":
  display-mode: normal
  plot:
  heatmap:
@@ -147,8 +147,9 @@ views:
  - -1
  - 0
  - 1
- regex('^b_(?!.*_se$)'):
- display-mode: normal
+ ?f"regex('signed_pi_value_(?!{params.primary_variable})')":
+ display-mode: hidden # Hide all signed pi cols that are not associated to the primary variable
+ ?f"regex('^b_(?!.*_se$)')":
  plot:
  heatmap:
  scale: linear
@@ -157,9 +158,9 @@ views:
  - "white"
  - "#6baed6"
  domain:
- - -1
+ - -10
  - 0
- - 1
+ - 10
  regex('^b_.*_se$'):
  display-mode: hidden
  ?for sample in params.samples:
@@ -242,7 +243,7 @@ views:
  display-mode: hidden
  regex('qval_.+'):
  display-mode: hidden
- regex('signed_pi_value_.+'):
+ ?f"regex('signed_pi_value_{params.primary_variable}.+')":
  display-mode: normal
  plot:
  heatmap:
@@ -255,8 +256,9 @@ views:
  - -1
  - 0
  - 1
- regex('b_.*(?<!_se(?!part2))'):
- display-mode: normal
+ ?f"regex('signed_pi_value_(?!{params.primary_variable})')":
+ display-mode: hidden # Hide all signed pi cols that are not associated to the primary variable
+ ?f"regex('^b_(?!.*_se$)')":
  plot:
  heatmap:
  scale: linear
@@ -265,9 +267,11 @@ views:
  - "white"
  - "#6baed6"
  domain:
- - -1
+ - -10
  - 0
- - 1
+ - 10
+ regex('^b_.*_se$'):
+ display-mode: hidden
  ?for sample in params.samples:
  ?sample:
  display-mode: hidden

workflow/resources/datavzrd/go-enrichment-template.yaml

@@ -52,7 +52,7 @@ views:
  }
  }
  spec-path: ?input.vega_bars
- vega-controls: true
+ vega-controls: false
  p_uncorrected:
  label: p-value
  plot:

workflow/resources/datavzrd/spia-template.yaml

@@ -59,7 +59,7 @@ views:
  }
  }
  spec-path: ?input.vega_circle
- vega-controls: true
+ vega-controls: false
  p-value for at least NDE genes:
  display-mode: detail
  total perturbation accumulation:
@@ -70,7 +70,10 @@ views:
  - "#B2182B"
  - "#F7F7F7"
  - "#2166AC"
- domain-mid: 0
+ domain: 
+ - -100
+ - 0
+ - 100
  p-value to observe a total accumulation:
  display-mode: detail
  Combined p-value:

workflow/scripts/compare_enrichment.py

@@ -185,7 +185,7 @@ def plot(df, effect_x, effect_y, title, xlabel, ylabel):
  )
  )
 
- point_selector = alt.selection_single(fields=["term"], empty="all")
+ point_selector = alt.selection_single(fields=["term"], empty=False)
  text_background = (
  alt.Chart(df)
  .mark_text(

workflow/scripts/compare_pathways.py

@@ -82,7 +82,7 @@ def prepare(df):
 
 alt.data_transformers.disable_max_rows()
 points = (
- alt.Chart(df)
+ alt.Chart(combined_pd)
  .mark_circle(size=15, tooltip={"content": "data"})
  .encode(
  alt.X(
@@ -102,8 +102,8 @@ def prepare(df):
  )
 )
 
-min_value = min(df[effect_x].min(), df[effect_y].min())
-max_value = max(df[effect_x].max(), df[effect_y].max())
+min_value = min(combined_pd[effect_x].min(), combined_pd[effect_y].min())
+max_value = max(combined_pd[effect_x].max(), combined_pd[effect_y].max())
 
 line = (
  alt.Chart(
@@ -140,7 +140,7 @@ def prepare(df):
 )
 
 text_background = (
- alt.Chart(df)
+ alt.Chart(combined_pd)
  .mark_text(
  align="left",
  baseline="middle",
@@ -158,7 +158,7 @@ def prepare(df):
 )
 
 text = (
- alt.Chart(df)
+ alt.Chart(combined_pd)
  .mark_text(
  align="left",
  baseline="middle",

workflow/scripts/postprocess_diffexp.py

@@ -20,8 +20,13 @@ def sort_columns(df, matching_columns):
  for prefix in matching_columns
  for suffix in ["_lower", "", "_upper", "_se"]
  ]
- other_columns = [col for col in df.columns if not col.startswith("b_")]
- return df[other_columns + b_column_order]
+ signed_pi_columns = [col for col in df.columns if col.startswith("signed_pi_value")]
+ other_columns = [
+ col
+ for col in df.columns
+ if not col.startswith("b_") and not col.startswith("signed_pi_value")
+ ]
+ return df[other_columns + b_column_order + signed_pi_columns]
 
 
 def sort_rows(df):