msreport 0.0.31__tar.gz → 0.0.32__tar.gz

{msreport-0.0.31 → msreport-0.0.32}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: msreport
-Version: 0.0.31
+Version: 0.0.32
 Summary: Post processing and analysis of quantitative proteomics data
 Author-email: "David M. Hollenstein" <hollenstein.david@gmail.com>
 License-Expression: Apache-2.0

{msreport-0.0.31 → msreport-0.0.32}/msreport/__init__.py

@@ -8,4 +8,4 @@ from msreport.fasta import import_protein_database
 from msreport.qtable import Qtable
 from msreport.reader import FragPipeReader, MaxQuantReader, SpectronautReader
 
-__version__ = "0.0.31"
+__version__ = "0.0.32"

{msreport-0.0.31 → msreport-0.0.32}/msreport/export.py

@@ -502,7 +502,7 @@ def _find_covered_region_boundaries(
     Examples:
         >>> coverage_mask = [True, True, False, False, True]
         >>> _find_covered_region_boundaries(coverage_mask)
-        ... [(0, 1), (4, 4)]
+        [(0, 1), (4, 4)]
     """
     start = []
     stop = []

{msreport-0.0.31 → msreport-0.0.32}/msreport/helper/maxlfq.py

@@ -113,9 +113,9 @@ def calculate_pairwise_mode_log_ratio_matrix(
         ...     ]
         ... )
         >>> calculate_pairwise_mode_log_ratio_matrix(array)
-        array([[ 0.       , -0.0849625, -1.       ],
-               [ 0.0849625,  0.       , -1.       ],
-               [ 1.       ,  1.       ,  0.       ]])
+        array([[ 0.        , -0.08496251, -1.       ],
+               [ 0.08496251,  0.       , -1.       ],
+               [ 1.        ,  1.       ,  0.       ]])
     """
     ratio_marix = _calculate_pairwise_centered_log_ratio_matrix(
         array, msreport.helper.mode, log_transformed=log_transformed

{msreport-0.0.31 → msreport-0.0.32}/msreport/plot/comparison.py

@@ -77,10 +77,15 @@ def volcano_ma(
         )
         special_entries = list(special_entries) + list(special_proteins)
 
-    data = qtable.get_data(exclude_invalid=exclude_invalid)
-    if annotation_column not in data.columns:
+    if annotation_column not in qtable.data.columns:
         annotation_column = qtable.id_column
 
+    data = qtable.get_data(exclude_invalid=exclude_invalid)
+    mask = np.ones(data.shape[0], dtype=bool)
+    for tag in [ratio_tag, expression_tag, pvalue_tag]:
+        mask = mask & np.isfinite(data[f"{tag} {comparison_group}"])
+    data = data[mask]
+
     scatter_size = 2 / (max(min(data.shape[0], 10000), 1000) / 1000)
 
     masks = {

{msreport-0.0.31 → msreport-0.0.32}/msreport/plot/multivariate.py

@@ -21,6 +21,7 @@ def sample_pca(
     pc_x: str = "PC1",
     pc_y: str = "PC2",
     exclude_invalid: bool = True,
+    exclude_missing: bool = False,
 ) -> tuple[plt.Figure, list[plt.Axes]]:
     """Figure to compare sample similarities with a principle component analysis.
 
@@ -44,11 +45,14 @@ def sample_pca(
             samples.
         exclude_invalid: If True, rows are filtered according to the Boolean entries of
             the "Valid" column.
+        exclude_missing: If True, only rows without any missing values are used.
 
     Returns:
         A matplotlib Figure and a list of Axes objects, containing the PCA plots.
     """
     design = qtable.get_design()
+    samples = qtable.get_samples()
+
     if design.shape[0] < 3:
         fig, ax = plt.subplots(1, 1, figsize=(2, 1.3))
         fig.suptitle(f'PCA of "{tag}" values', y=1.1)
@@ -65,13 +69,22 @@ def sample_pca(
         return fig, np.array([ax])
 
     table = qtable.make_sample_table(
-        tag, samples_as_columns=True, exclude_invalid=exclude_invalid
+        tag, samples_as_columns=True, exclude_invalid=False
     )
+
+    inclusion_mask = np.ones(qtable.data.shape[0], dtype=bool)
+    if exclude_invalid:
+        inclusion_mask = inclusion_mask & qtable["Valid"]
+    if exclude_missing:
+        _non_missing_masks = [(qtable[f"Missing {s}"] == 0) for s in samples]
+        inclusion_mask = inclusion_mask & (np.all(_non_missing_masks, axis=0))
+    table = table[inclusion_mask]
+
     table = table.replace({0: np.nan})
     table = table[np.isfinite(table).sum(axis=1) > 0]
     if not msreport.helper.intensities_in_logspace(table):
         table = np.log2(table)
-    table[table.isna()] = 0
+    table = table.fillna(0)
 
     table = table.transpose()
     sample_index = table.index.tolist()
@@ -203,6 +216,7 @@ def sample_pca(
 def expression_clustermap(
     qtable: Qtable,
     exclude_invalid: bool = True,
+    exclude_missing: bool = False,
     remove_imputation: bool = True,
     mean_center: bool = False,
     cluster_samples: bool = True,
@@ -218,6 +232,7 @@ def expression_clustermap(
         qtable: A `Qtable` instance, which data is used for plotting.
         exclude_invalid: If True, rows are filtered according to the Boolean entries of
             the "Valid" column.
+        exclude_missing: If True, only rows without any missing values are used.
         remove_imputation: If True, imputed values are set to 0 before clustering.
             Defaults to True.
         mean_center: If True, the data is mean-centered before clustering. Defaults to
@@ -242,25 +257,29 @@ def expression_clustermap(
     if len(samples) < 2:
         raise ValueError("At least two samples are required to generate a clustermap.")
 
-    data = qtable.make_expression_table(samples_as_columns=True)
+    data = qtable.make_expression_table(samples_as_columns=True, exclude_invalid=False)
     data = data[samples]
+    data = data.fillna(0)
 
-    for sample in samples:
-        if remove_imputation:
-            data.loc[qtable.data[f"Missing {sample}"], sample] = 0
-        data[sample] = data[sample].fillna(0)
-
-    if not mean_center:
-        # Hide missing values in the heatmap, making them appear white
-        mask_values = qtable.data[
+    if not mean_center:  # Hide missing values in the heatmap, making them appear white
+        hide_values_mask = qtable.data[
             [f"Missing {sample}" for sample in samples]
         ].to_numpy()
     else:
-        mask_values = np.zeros(data.shape, dtype=bool)
+        hide_values_mask = np.zeros(data.shape, dtype=bool)
+
+    if remove_imputation:
+        for sample in samples:
+            data.loc[qtable.data[f"Missing {sample}"], sample] = 0
 
+    inclusion_mask = np.ones(data.shape[0], dtype=bool)
     if exclude_invalid:
-        data = data[qtable.data["Valid"]]
-        mask_values = mask_values[qtable.data["Valid"]]
+        inclusion_mask = inclusion_mask & qtable["Valid"]
+    if exclude_missing:
+        _non_missing_masks = [(qtable[f"Missing {s}"] == 0) for s in samples]
+        inclusion_mask = inclusion_mask & (np.all(_non_missing_masks, axis=0))
+    hide_values_mask = hide_values_mask[inclusion_mask]
+    data = data[inclusion_mask]
 
     color_wheel = ColorWheelDict()
     for exp in experiments:
@@ -314,7 +333,7 @@ def expression_clustermap(
         col_cluster=cluster_samples,
         col_colors=sample_colors,
         row_colors=["#000000" for _ in range(len(data))],
-        mask=mask_values,
+        mask=hide_values_mask,
         method=cluster_method,
         metric="euclidean",
         **heatmap_args,

{msreport-0.0.31 → msreport-0.0.32}/msreport/reader.py

@@ -545,7 +545,12 @@ class FragPipeReader(ResultReader):
     """FragPipe result reader.
 
     Methods:
-        import_design: Reads a "fragpipe-files.fp-manifest" file and returns a
+        import_design: Depending on the quantification strategy, imports either the
+            manifest file or the experiment annotation file and returns a processed
+            design dataframe.
+        import_manifest: Reads a "fragpipe-files.fp-manifest" file and returns a
+            processed design dataframe.
+        import_experiment_annotation: Reads a "experiment_annotation" file and returns a
             processed design dataframe.
         import_proteins: Reads a "combined_protein.tsv" or "protein.tsv" file and
             returns a processed dataframe, conforming to the MsReport naming
@@ -589,12 +594,8 @@ class FragPipeReader(ResultReader):
         "ions": "combined_ion.tsv",
         "ion_evidence": "ion.tsv",
         "psm_evidence": "psm.tsv",
-        "design": "fragpipe-files.fp-manifest",
-    }
-    isobar_filenames: dict[str, str] = {
-        "proteins": "protein.tsv",
-        "peptides": "peptide.tsv",
-        "ions": "ion.tsv",
+        "manifest": "fragpipe-files.fp-manifest",
+        "experiment_annotation": "experiment_annotation.tsv",
     }
     sil_filenames: dict[str, str] = {
         "proteins": "combined_protein_label_quant.tsv",
@@ -675,14 +676,27 @@ class FragPipeReader(ResultReader):
         self._isobar: bool = isobar
         self._sil: bool = sil
         self._contaminant_tag: str = contaminant_tag
-        if isobar:
-            self.filenames = self.isobar_filenames
-        elif sil:
-            self.filenames = self.sil_filenames
+
+        self.filenames = self.default_filenames.copy()
+        if sil:
+            self.filenames.update(self.sil_filenames)
+
+    def import_design(self, sort: bool = False) -> pd.DataFrame:
+        """Reads the experimental design file and returns a processed design dataframe.
+
+        Depending on the quantification strategy (isobaric or label-free/sil), either
+        the experiment annotation file or the manifest file is imported.
+
+        Args:
+            sort: If True, the design dataframe is sorted by "Experiment" and
+                "Replicate"; default False.
+        """
+        if self._isobar:
+            return self.import_experiment_annotation(sort=sort)
         else:
-            self.filenames = self.default_filenames
+            return self.import_manifest(sort=sort)
 
-    def import_design(
+    def import_manifest(
         self, filename: Optional[str] = None, sort: bool = False
     ) -> pd.DataFrame:
         """Read a 'fp-manifest' file and returns a processed design dataframe.
@@ -709,7 +723,7 @@ class FragPipeReader(ResultReader):
             FileNotFoundError: If the specified manifest file does not exist.
         """
         if filename is None:
-            filepath = os.path.join(self.data_directory, self.filenames["design"])
+            filepath = os.path.join(self.data_directory, self.filenames["manifest"])
         else:
             filepath = os.path.join(self.data_directory, filename)
         if not os.path.exists(filepath):
@@ -748,6 +762,63 @@ class FragPipeReader(ResultReader):
             design.reset_index(drop=True, inplace=True)
         return design
 
+    def import_experiment_annotation(
+        self, filename: Optional[str] = None, sort: bool = False
+    ) -> pd.DataFrame:
+        """Read a 'experiment_annotation' file and returns a processed design dataframe.
+
+        The annotation columns "sample", "channel", and "plex" are mapped to the design
+        table columns "Sample", "Channel", and "Plex". The "Experiment" and "Replicate"
+        columns are extracted from the "Sample" column by splitting at the last
+        underscore, if there is no underscore, "Replicate" is set to an empty string.
+
+        Note that this convention of splitting the "Sample" column does confirm to the
+        FragPipe convention, but FragPipe does not enforce it for the experiment
+        annotation file.
+
+        Args:
+            filename: Allows specifying an alternative filename, otherwise the default
+                filename is used.
+            sort: If True, the design dataframe is sorted by "Experiment" and
+                "Replicate"; default False.
+
+        Returns:
+            A dataframe containing the processed design table with columns:
+            "Sample", "Experiment", "Replicate", "Channel", and "Plex".
+
+        Raises:
+            FileNotFoundError: If the specified manifest file does not exist.
+        """
+        if filename is None:
+            filepath = os.path.join(
+                self.data_directory, self.filenames["experiment_annotation"]
+            )
+        else:
+            filepath = os.path.join(self.data_directory, filename)
+        if not os.path.exists(filepath):
+            raise FileNotFoundError(
+                f"File '{filepath}' does not exist. Please check the file path."
+            )
+
+        annotation = pd.read_csv(filepath, sep="\t")
+
+        design = pd.DataFrame(
+            {
+                "Sample": annotation["sample"],
+                "Experiment": annotation["sample"].str.rsplit("_", n=1).str[0],
+                "Replicate": annotation["sample"].str.rsplit("_", n=1).str[1],
+                "Channel": annotation["channel"],
+                "Plex": annotation["plex"],
+            }
+        )
+        design["Replicate"] = design["Replicate"].fillna("")
+
+        if sort:
+            design.sort_values(by=["Experiment", "Replicate"], inplace=True)
+            design.reset_index(drop=True, inplace=True)
+
+        return design
+
     def import_proteins(
         self,
         filename: Optional[str] = None,
@@ -1034,6 +1105,7 @@ class FragPipeReader(ResultReader):
             )
             df["Modified sequence"] = mod_entries["Modified sequence"]
             df["Modifications"] = mod_entries["Modifications"]
+            df = self._add_modification_localization_string_to_psm_evidence(df)
         return df
 
     def _add_protein_entries(self, df: pd.DataFrame) -> pd.DataFrame:
@@ -1207,6 +1279,66 @@ class FragPipeReader(ResultReader):
             new_df[new_column] = localization_strings
         return new_df
 
+    def _add_modification_localization_string_to_psm_evidence(
+        self, df: pd.DataFrame
+    ) -> pd.DataFrame:
+        """Adds a modification localization string column to a PSM evidence table.
+
+        Extracts localization probabilities from all columns in the form
+        f"{aa:modification}", converts them into the standardized modification
+        localization string format used by msreport, and adds a new column
+        "Modification localization string".
+
+        Probabilities are written in the format
+        "Mod1@Site1:Probability1,Site2:Probability2;Mod2@Site3:Probability3",
+        e.g. "15.9949@11:1.000;79.9663@3:0.200,4:0.800". Refer to
+        `msreport.peptidoform.make_localization_string` for details.
+
+        Args:
+            df: A dataframe containing PSM tables from FragPipe.
+
+        Returns:
+            A copy of the input dataframe with the added
+            "Modification localization string" column.
+        """
+        new_df = df.copy()
+        _search_tag = " Best Localization"
+        mod_localization_columns = [
+            c.strip(_search_tag) for c in new_df.columns if c.endswith(_search_tag)
+        ]
+        if not mod_localization_columns:
+            new_df["Modification localization string"] = ""
+            return new_df
+
+        df[mod_localization_columns] = (
+            df[mod_localization_columns].astype(str).replace("nan", "")
+        )
+        row_mod_probabilities: list[dict[str, dict[int, float]]] = [
+            {} for i in range(df.shape[0])
+        ]
+        for mod_localization_column in mod_localization_columns:
+            modification = mod_localization_column.split(":")[1]
+            for modification_probabilities, probability_sequence in zip(
+                row_mod_probabilities, df[mod_localization_column]
+            ):
+                if not probability_sequence:
+                    continue
+                _, probabilities = msreport.peptidoform.parse_modified_sequence(
+                    probability_sequence, "(", ")"
+                )
+                modification_probabilities[modification] = {
+                    site: float(probability) for site, probability in probabilities
+                }
+
+        localization_strings = []
+        for localization_probabilities in row_mod_probabilities:
+            localization_string = msreport.peptidoform.make_localization_string(
+                localization_probabilities
+            )
+            localization_strings.append(localization_string)
+        new_df["Modification localization string"] = localization_strings
+        return new_df
+
 
 class SpectronautReader(ResultReader):
     """Spectronaut result reader.

{msreport-0.0.31 → msreport-0.0.32}/msreport.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: msreport
-Version: 0.0.31
+Version: 0.0.32
 Summary: Post processing and analysis of quantitative proteomics data
 Author-email: "David M. Hollenstein" <hollenstein.david@gmail.com>
 License-Expression: Apache-2.0

{msreport-0.0.31 → msreport-0.0.32}/pyproject.toml

@@ -112,3 +112,11 @@ module = [
   "yaml.*",
 ]
 follow_untyped_imports = true
+
+[tool.pytest.ini_options]
+addopts = "--doctest-modules"
+doctest_optionflags = "NORMALIZE_WHITESPACE IGNORE_EXCEPTION_DETAIL"
+testpaths = [
+    "tests",
+    "msreport",
+]

{msreport-0.0.31 → msreport-0.0.32}/tests/test_plot.py

@@ -1,3 +1,4 @@
+import matplotlib
 import numpy as np
 import pandas as pd
 import pytest
@@ -5,6 +6,9 @@ import pytest
 import msreport.plot
 import msreport.qtable
 
+# Use the 'Agg' backend for plotting tests to prevent TclError in headless environments.
+matplotlib.use("Agg")
+
 
 @pytest.fixture
 def example_data():