Rethink grouping in Box Plots

protzilla/data_preprocessing/imputation.py

@@ -359,6 +359,7 @@ def by_knn_plot(
     graph_type_quantities,
     group_by,
     visual_transformation,
+    proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
@@ -367,6 +368,7 @@ def by_knn_plot(
         graph_type_quantities,
         group_by,
         visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -377,6 +379,7 @@ def by_normal_distribution_sampling_plot(
     graph_type_quantities,
     group_by,
     visual_transformation,
+    proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
@@ -385,6 +388,7 @@ def by_normal_distribution_sampling_plot(
         graph_type_quantities,
         group_by,
         visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -395,6 +399,7 @@ def by_simple_imputer_plot(
     graph_type_quantities,
     group_by,
     visual_transformation,
+    proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
@@ -403,6 +408,7 @@ def by_simple_imputer_plot(
         graph_type_quantities,
         group_by,
         visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -413,6 +419,7 @@ def by_min_per_sample_plot(
     graph_type_quantities,
     group_by,
     visual_transformation,
+    proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
@@ -421,6 +428,7 @@ def by_min_per_sample_plot(
         graph_type_quantities,
         group_by,
         visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -431,6 +439,7 @@ def by_min_per_protein_plot(
     graph_type_quantities,
     group_by,
     visual_transformation,
+    proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
@@ -439,6 +448,7 @@ def by_min_per_protein_plot(
         graph_type_quantities,
         group_by,
         visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -449,6 +459,7 @@ def by_min_per_dataset_plot(
     graph_type_quantities,
     group_by,
     visual_transformation,
+    proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
@@ -457,6 +468,7 @@ def by_min_per_dataset_plot(
         graph_type_quantities,
         group_by,
         visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -471,6 +483,7 @@ def _build_box_hist_plot(
     graph_type_quantities: str = "Pie chart",
     group_by: str = "None",
     visual_transformation: str = "linear",
+    proteins_of_interest= None,
 ) -> list[Figure]:
     """
     This function creates two visualisations:
@@ -506,6 +519,7 @@ def _build_box_hist_plot(
             heading="Distribution of Protein Intensities",
             group_by=group_by,
             visual_transformation=visual_transformation,
+            proteins_of_interest=proteins_of_interest,
             y_title="Intensity",
         )
     elif graph_type == "Histogram":

protzilla/data_preprocessing/normalisation.py

@@ -233,14 +233,16 @@ def by_z_score_plot(
         method_outputs,
         graph_type,
         group_by,
-        visual_transformation
+        visual_transformation,
+        proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
         method_outputs["protein_df"],
         graph_type,
         group_by,
-        visual_transformation
+        visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -249,13 +251,15 @@ def by_median_plot(
         method_outputs,
         graph_type,
         group_by,
-        visual_transformation
+        visual_transformation,
+        proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
         method_outputs["protein_df"],
         graph_type, group_by,
-        visual_transformation
+        visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -264,13 +268,15 @@ def by_totalsum_plot(
         method_outputs,
         graph_type,
         group_by,
-        visual_transformation
+        visual_transformation,
+        proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
         method_outputs["protein_df"],
         graph_type, group_by,
-        visual_transformation
+        visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
@@ -279,18 +285,20 @@ def by_reference_protein_plot(
         method_outputs,
         graph_type,
         group_by,
-        visual_transformation
+        visual_transformation,
+        proteins_of_interest=None,
 ):
     return _build_box_hist_plot(
         method_inputs["protein_df"],
         method_outputs["protein_df"],
         graph_type,
         group_by,
-        visual_transformation
+        visual_transformation,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
 
-def _build_box_hist_plot(df, result_df, graph_type, group_by, visual_transformation):
+def _build_box_hist_plot(df, result_df, graph_type, group_by, visual_transformation, proteins_of_interest=None):
     if graph_type == "Boxplot":
         fig = create_box_plots(
             dataframe_a=df,
@@ -302,6 +310,7 @@ def _build_box_hist_plot(df, result_df, graph_type, group_by, visual_transformat
             y_title="Intensity",
             group_by=group_by,
             visual_transformation=visual_transformation,
+            proteins_of_interest=proteins_of_interest,
         )
     if graph_type == "Histogram":
         fig = create_histograms(

protzilla/data_preprocessing/plots.py

@@ -55,7 +55,7 @@ def create_bar_plot(
     names_of_sectors: "list[str]",
     values_of_sectors: "list[int]",
     heading: str = "",
-    colour: "list[str]" = PROTZILLA_DISCRETE_COLOR_OUTLIER_SEQUENCE,
+    color: "list[str]" = PROTZILLA_DISCRETE_COLOR_OUTLIER_SEQUENCE,
     y_title: str = "",
     x_title: str = "",
 ) -> Figure:
@@ -76,7 +76,7 @@ def create_bar_plot(
     fig = px.bar(
         x=names_of_sectors,
         y=values_of_sectors,
-        color=colour[: len(values_of_sectors)],
+        color=color[: len(values_of_sectors)],
         color_discrete_map="identity",
     )
 
@@ -110,6 +110,7 @@ def create_box_plots(
     y_title: str = "",
     x_title: str = "",
     group_by: str = "None",
+    proteins_of_interest=None,
     visual_transformation: str = "linear",
 ) -> Figure:
     """
@@ -129,6 +130,7 @@ def create_box_plots(
     :param y_title: Optional y-axis title for graphs.
     :param x_title: Optional x-axis title for graphs.
     :param group_by: Optional argument to create a grouped boxplot\
+    :param proteins_of_interest: List of proteins to be included in the boxplot
     :param visual_transformation: Visual transformation of the y-axis data.
     graph. Arguments can be either "Sample" to group by sample or\
     "Protein ID" to group by protein. Leave "None" to get ungrouped\
@@ -141,19 +143,29 @@ def create_box_plots(
             f"""Group_by parameter  must be "None" or
                 "Sample" or "Protein ID" but is {group_by}"""
         )
-    intensity_name_a = default_intensity_column(dataframe_a)
-    intensity_name_b = default_intensity_column(dataframe_b)
+    if len(proteins_of_interest) == 0:
+        proteins_of_interest = dataframe_a["Protein ID"]
+    elif isinstance(proteins_of_interest, str):
+        proteins_of_interest = [proteins_of_interest]
+
+    filtered_df = dataframe_a[dataframe_a["Protein ID"].isin(proteins_of_interest)]
+    filtered_result_df = dataframe_b[
+        dataframe_b["Protein ID"].isin(proteins_of_interest)
+    ]
+    intensity_name_a = default_intensity_column(filtered_df)
+    intensity_name_b = default_intensity_column(filtered_result_df)
+
     if group_by in {"Sample", "Protein ID"}:
         fig = make_subplots(rows=1, cols=2)
         trace0 = go.Box(
-            y=dataframe_a[intensity_name_a],
-            x=dataframe_a[group_by],
+            y=filtered_df[intensity_name_a],
+            x=filtered_df[group_by],
             marker_color=PROTZILLA_DISCRETE_COLOR_OUTLIER_SEQUENCE[0],
             name=name_a,
         )
         trace1 = go.Box(
-            y=dataframe_b[intensity_name_b],
-            x=dataframe_b[group_by],
+            y=filtered_result_df[intensity_name_b],
+            x=filtered_result_df[group_by],
             marker_color=PROTZILLA_DISCRETE_COLOR_OUTLIER_SEQUENCE[1],
             name=name_b,
         )
@@ -164,12 +176,12 @@ def create_box_plots(
     elif group_by == "None":
         fig = make_subplots(rows=1, cols=2)
         trace0 = go.Box(
-            y=dataframe_a[intensity_name_a],
+            y=filtered_df[intensity_name_a],
             marker_color=PROTZILLA_DISCRETE_COLOR_OUTLIER_SEQUENCE[0],
             name=name_a,
         )
         trace1 = go.Box(
-            y=dataframe_b[intensity_name_b],
+            y=filtered_result_df[intensity_name_b],
             marker_color=PROTZILLA_DISCRETE_COLOR_OUTLIER_SEQUENCE[1],
             name=name_b,
         )
@@ -208,6 +220,7 @@ def create_histograms(
     heading: str = "",
     y_title: str = "",
     x_title: str = "",
+    proteins_of_interest=None,
     visual_transformation: str = "linear",
     overlay: bool = False,
 ) -> Figure:

protzilla/data_preprocessing/transformation.py

@@ -44,13 +44,28 @@ def by_log(protein_df: pd.DataFrame, peptide_df: pd.DataFrame | None, log_base="
     return dict(protein_df=transformed_df, peptide_df=transformed_peptide_df)
 
 
-def by_log_plot(method_inputs, method_outputs, graph_type, group_by):
+def by_log_plot(
+    method_inputs,
+    method_outputs,
+    graph_type,
+    group_by,
+    proteins_of_interest=None,
+):
     return _build_box_hist_plot(
-        method_inputs["protein_df"], method_outputs["protein_df"], graph_type, group_by
+        method_inputs["protein_df"],
+        method_outputs["protein_df"],
+        graph_type,
+        group_by,
+        [] if proteins_of_interest is None else proteins_of_interest,
     )
 
-
-def _build_box_hist_plot(df, result_df, graph_type, group_by):
+def _build_box_hist_plot(
+    df,
+    result_df,
+    graph_type,
+    group_by,
+    proteins_of_interest=None,
+):
     if graph_type == "Boxplot":
         fig = create_box_plots(
             dataframe_a=df,
@@ -59,6 +74,7 @@ def _build_box_hist_plot(df, result_df, graph_type, group_by):
             name_b="After Transformation",
             heading="Distribution of Protein Intensities",
             group_by=group_by,
+            proteins_of_interest=proteins_of_interest,
             y_title="Intensity",
         )
     if graph_type == "Histogram":
@@ -68,6 +84,7 @@ def _build_box_hist_plot(df, result_df, graph_type, group_by):
             name_a="Before Transformation",
             name_b="After Transformation",
             heading="Distribution of Protein Intensities",
+            proteins_of_interest=proteins_of_interest,
             x_title="Protein Intensities",
             y_title="Frequency of Protein Intensities",
         )

protzilla/steps.py

@@ -661,7 +661,7 @@ def _clear_future_steps(self, index: int | None = None) -> None:
             index = self.current_step_index
         if index == len(self.all_steps) - 1:
             return
-        for step in self.all_steps[index + 1 :]:
+        for step in self.all_steps[index:]:
             step.output = Output()
             step.messages = Messages()
             step.plots = Plots()

tests/protzilla/data_preprocessing/test_imputation.py

@@ -165,6 +165,7 @@ def test_imputation_min_value_per_df(
         "Bar chart",
         "Sample",
         "linear",
+        [],
     )
     if show_figures:
         fig1.show()
@@ -200,6 +201,7 @@ def test_imputation_min_value_per_sample(
         "Bar chart",
         "Sample",
         "linear",
+        [],
     )
     if show_figures:
         fig1.show()
@@ -235,6 +237,7 @@ def test_imputation_min_value_per_protein(
         "Bar chart",
         "Sample",
         "linear",
+        [],
     )
     if show_figures:
         fig1.show()
@@ -270,6 +273,7 @@ def test_imputation_mean_per_protein(
         "Bar chart",
         "Sample",
         "linear",
+        [],
     )
     if show_figures:
         fig1.show()
@@ -303,6 +307,7 @@ def test_imputation_knn(show_figures, input_imputation_df, assertion_df_knn):
         "Bar chart",
         "Sample",
         "linear",
+        [],
     )
     if show_figures:
         fig1.show()
@@ -345,6 +350,7 @@ def test_imputation_normal_distribution_sampling(show_figures, input_imputation_
         "Bar chart",
         "Sample",
         "linear",
+        [],
     )
     if show_figures:
         fig1.show()