Merge pull request #1648 from nilchia/fix_header_flexy

change flexynesis input/output from csv to tabular

Author: bgruening

tools/flexynesis/convert.py

@@ -0,0 +1,27 @@
+#!/usr/bin/env python
+
+import sys
+
+import pandas as pd
+
+
+def tabular_to_csv(tabular_file, csv_file):
+    """Convert tabular (TSV) to CSV"""
+    data = pd.read_csv(tabular_file, sep="\t")
+    data.to_csv(csv_file, index=False)
+
+
+def csv_to_tabular(csv_file, tabular_file):
+    """Convert CSV to tabular (TSV)"""
+    data = pd.read_csv(csv_file)
+    data.to_csv(tabular_file, sep="\t", index=False)
+
+
+if __name__ == "__main__":
+    input_file = sys.argv[1]
+    output_file = sys.argv[2]
+
+    if input_file.endswith('.csv'):
+        csv_to_tabular(input_file, output_file)
+    else:
+        tabular_to_csv(input_file, output_file)

tools/flexynesis/flexynesis.xml

@@ -214,21 +214,25 @@ def plot_boxplot(categorical_x, numerical_y, title_x='Categories', title_y='Valu
 def generate_dimred_plots(embeddings, matched_labels, args, output_dir, output_name_base):
     """Generate dimensionality reduction plots"""
 
-    # Parse target variables
-    target_vars = [var.strip() for var in args.target_variables.split(',')]
+    # Parse target values from comma-separated string
+    if args.target_value:
+        target_values = [val.strip() for val in args.target_value.split(',')]
+    else:
+        # If no target values specified, use all unique variables
+        target_values = matched_labels['variable'].unique().tolist()
 
-    print(f"Generating {args.method.upper()} plots for {len(target_vars)} target variable(s): {', '.join(target_vars)}")
+    print(f"Generating {args.method.upper()} plots for {len(target_values)} target variable(s): {', '.join(target_values)}")
 
     # Check variables
     available_vars = matched_labels['variable'].unique()
-    missing_vars = [var for var in target_vars if var not in available_vars]
+    missing_vars = [var for var in target_values if var not in available_vars]
 
     if missing_vars:
         print(f"Warning: The following target variables were not found in the data: {', '.join(missing_vars)}")
         print(f"Available variables: {', '.join(available_vars)}")
 
     # Filter to only process available variables
-    valid_vars = [var for var in target_vars if var in available_vars]
+    valid_vars = [var for var in target_values if var in available_vars]
 
     if not valid_vars:
         raise ValueError(f"None of the specified target variables were found in the data. Available: {', '.join(available_vars)}")
@@ -989,8 +993,6 @@ def main():
                         help="Path to input data embeddings file (CSV or tabular format). Required for dimred plots.")
     parser.add_argument("--method", type=str, default='pca', choices=['pca', 'umap'],
                         help="Transformation method ('pca' or 'umap'). Default is 'pca'. Used for dimred plots.")
-    parser.add_argument("--target_variables", type=str, required=False,
-                        help="Comma-separated list of target variables to plot.")
 
     # Arguments for Kaplan-Meier
     parser.add_argument("--survival_data", type=str,
@@ -1024,7 +1026,7 @@ def main():
     parser.add_argument("--random_state", type=int, default=42,
                         help="Random seed for reproducibility. Default is 42")
 
-    # Arguments for scatter plot, heatmap, PR curves, ROC curves, and box plots
+    # Arguments for dimred, scatter plot, heatmap, PR curves, ROC curves, and box plots
     parser.add_argument("--target_value", type=str, default=None,
                         help="Target value for scatter plot.")
 
@@ -1057,8 +1059,6 @@ def main():
                 raise ValueError("--labels is required for dimensionality reduction plots")
             if not args.method:
                 raise ValueError("--method is required for dimensionality reduction plots")
-            if not args.target_variables:
-                raise ValueError("--target_variables is required for dimensionality reduction plots")
 
         if args.plot_type in ['kaplan_meier']:
             if not args.survival_data:

tools/flexynesis/flexynesis_plot.py

@@ -152,7 +152,7 @@ for class_label in classes:
             <when value="dimred">
                 <expand macro="plots_common_param">
                     <expand macro="plots_common_input"/>
-                    <param argument="--embeddings" type="data" format="tabular,csv" label="Embeddings" help="Generated by flexynesis"/>
+                    <param argument="--embeddings" type="data" format="tabular" label="Embeddings" help="Generated by flexynesis"/>
                     <param argument="--label" type="data_column" data_ref="labels" label="Column in the labels file to use for coloring the points in the plot"/>
                     <param name="method" type="select" label="Transformation method">
                         <option value="pca" selected="true">PCA</option>
@@ -206,10 +206,10 @@ for class_label in classes:
             <param name="non_commercial_use" value="True"/>
             <conditional name="plot_conditional">
                 <param name="plot_type" value="dimred"/>
-                <param name="embeddings" value="embeddings.csv"/>
+                <param name="embeddings" value="embeddings.tabular"/>
                 <param name="label" value="6"/>
                 <param name="method" value="pca"/>
-                <param name="labels" value="labels.csv"/>
+                <param name="labels" value="labels.tabular"/>
                 <param name="format" value="jpg"/>
                 <param name="dpi" value="300"/>
             </conditional>
@@ -227,7 +227,7 @@ for class_label in classes:
             <param name="non_commercial_use" value="True"/>
             <conditional name="plot_conditional">
                 <param name="plot_type" value="scatter"/>
-                <param name="labels" value="labels_scatter.csv"/>
+                <param name="labels" value="labels_scatter.tabular"/>
                 <param name="true_label" value="5"/>
                 <param name="predicted_label" value="6"/>
                 <param name="format" value="jpg"/>
@@ -247,7 +247,7 @@ for class_label in classes:
             <param name="non_commercial_use" value="True"/>
             <conditional name="plot_conditional">
                 <param name="plot_type" value="concordance_heatmap"/>
-                <param name="labels" value="labels.csv"/>
+                <param name="labels" value="labels.tabular"/>
                 <param name="true_label" value="5"/>
                 <param name="predicted_label" value="6"/>
                 <param name="format" value="jpg"/>
@@ -267,7 +267,7 @@ for class_label in classes:
             <param name="non_commercial_use" value="True"/>
             <conditional name="plot_conditional">
                 <param name="plot_type" value="pr_curve"/>
-                <param name="labels" value="labels_pr.csv"/>
+                <param name="labels" value="labels_pr.tabular"/>
                 <param name="format" value="jpg"/>
                 <param name="dpi" value="300"/>
             </conditional>
@@ -285,7 +285,7 @@ for class_label in classes:
             <param name="non_commercial_use" value="True"/>
             <conditional name="plot_conditional">
                 <param name="plot_type" value="roc_curve"/>
-                <param name="labels" value="labels_pr.csv"/>
+                <param name="labels" value="labels_pr.tabular"/>
                 <param name="format" value="jpg"/>
                 <param name="dpi" value="300"/>
             </conditional>
@@ -303,7 +303,7 @@ for class_label in classes:
             <param name="non_commercial_use" value="True"/>
             <conditional name="plot_conditional">
                 <param name="plot_type" value="box_plot"/>
-                <param name="labels" value="labels_pr.csv"/>
+                <param name="labels" value="labels_pr.tabular"/>
                 <param name="format" value="jpg"/>
                 <param name="dpi" value="300"/>
             </conditional>

tools/flexynesis/flexynesis_plot.xml

@@ -156,8 +156,8 @@ def split_and_save_data(data, ratio=0.7, output_dir='.'):
     # Save train and test data
     for key in data.keys():
         try:
-            train_data[key].to_csv(os.path.join(output_dir, 'train', f'{key}.csv'))
-            test_data[key].to_csv(os.path.join(output_dir, 'test', f'{key}.csv'))
+            train_data[key].to_csv(os.path.join(output_dir, 'train', f'{key}.tabular'), sep='\t')
+            test_data[key].to_csv(os.path.join(output_dir, 'test', f'{key}.tabular'), sep='\t')
         except Exception as e:
             print(f"Error saving {key}: {e}")
             continue
@@ -244,8 +244,8 @@ def main():
 
             binarized_matrix = binarize_mutations(mutations_df, gene_idx=args.gene_idx, sample_idx=args.sample_idx)
             # Save binarized matrix
-            output_file = os.path.join(args.out, 'binarized_mutations.csv')
-            binarized_matrix.to_csv(output_file)
+            output_file = os.path.join(args.out, 'binarized_mutations.tabular')
+            binarized_matrix.to_csv(output_file, sep='\t')
             print(f"Binarized mutation matrix saved to {output_file}")
 
     except Exception as e: