Merge pull request #209 from singjc/master

Add exclude_decoys option and filtering in BaseWriter

Author: singjc

pyprophet/_config.py

@@ -716,4 +716,7 @@ class ExportIOConfig(BaseIOConfig):
     intensity_calibration: bool = True
     min_fragments: int = 4
     keep_decoys: bool = False  # Whether to keep decoy entries in the library
-    rt_unit: Literal["iRT", "RT"] = "iRT"
+    rt_unit: Literal["iRT", "RT"] = "iRT"
+
+    # TSV/Matrix export options
+    exclude_decoys: bool = True  # Whether to exclude decoy entries from TSV/matrix export (default: True, exclude decoys)

pyprophet/cli/export.py

@@ -159,6 +159,13 @@ def export():
     type=float,
     help="[format: matrix/legacy] Maximum PEP to consider for good alignments when use_alignment is enabled.",
 )
+@click.option(
+    "--exclude-decoys/--no-exclude-decoys",
+    "exclude_decoys",
+    default=True,
+    show_default=True,
+    help="Exclude decoy entries from the exported results. Use --no-exclude-decoys to retain decoys.",
+)
 @measure_memory_usage_and_time
 def export_tsv(
     infile,
@@ -176,6 +183,7 @@ def export_tsv(
     max_global_protein_qvalue,
     use_alignment,
     max_alignment_pep,
+    exclude_decoys,
 ):
     """
     Export Proteomics/Peptidoform TSV/CSV tables
@@ -207,6 +215,7 @@ def export_tsv(
         max_global_protein_qvalue=max_global_protein_qvalue,
         use_alignment=use_alignment,
         max_alignment_pep=max_alignment_pep,
+        exclude_decoys=exclude_decoys,
     )
 
     reader = ReaderDispatcher.get_reader(config)
@@ -329,6 +338,13 @@ def export_tsv(
     type=click.Choice(["none", "median", "medianmedian", "quantile"]),
     help="[format: matrix/legacy] Normalization method to apply to the quantification matrix.",
 )
+@click.option(
+    "--exclude-decoys/--no-exclude-decoys",
+    "exclude_decoys",
+    default=True,
+    show_default=True,
+    help="Exclude decoy entries from the exported matrix. Use --no-exclude-decoys to retain decoys.",
+)
 @measure_memory_usage_and_time
 def export_matrix(
     infile,
@@ -347,6 +363,7 @@ def export_matrix(
     top_n,
     consistent_top,
     normalization,
+    exclude_decoys,
 ):
     """
     Export Proteomics/Peptidoform Quantification Matrix
@@ -381,6 +398,7 @@ def export_matrix(
         top_n=top_n,
         consistent_top=consistent_top,
         normalization=normalization,
+        exclude_decoys=exclude_decoys,
     )
 
     reader = ReaderDispatcher.get_reader(config)

pyprophet/io/_base.py

@@ -589,6 +589,14 @@ def export_results(self, data: pd.DataFrame):
         """
         cfg = self.config
 
+        # Filter out decoys if exclude_decoys is True
+        if cfg.exclude_decoys and "decoy" in data.columns:
+            initial_count = len(data)
+            data = data.loc[data["decoy"].eq(0)].copy()
+            decoy_count = initial_count - len(data)
+            if decoy_count > 0:
+                logger.info(f"Excluded {decoy_count} decoy entries from export.")
+
         sep = "," if cfg.out_type == "csv" else "\t"
 
         if cfg.export_format == "legacy_split":
@@ -737,6 +745,14 @@ def export_quant_matrix(self, data: pd.DataFrame) -> pd.DataFrame:
         """
         cfg = self.config
 
+        # Filter out decoys if exclude_decoys is True
+        if cfg.exclude_decoys and "decoy" in data.columns:
+            initial_count = len(data)
+            data = data[data["decoy"] == 0]
+            decoy_count = initial_count - len(data)
+            if decoy_count > 0:
+                logger.info(f"Excluded {decoy_count} decoy entries from quantification matrix.")
+
         # Check if data is empty
         if data.empty:
             raise ValueError(