protdata 0.1.4__tar.gz → 0.1.5__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,7 +1,6 @@
1
1
  BSD 3-Clause License
2
2
 
3
- Copyright (c) 2025, Chan Zuckerberg Biohub
4
-
3
+ Copyright (c) 2026, CZ Biohub SF, LLC
5
4
  Redistribution and use in source and binary forms, with or without
6
5
  modification, are permitted provided that the following conditions are met:
7
6
 
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: protdata
3
- Version: 0.1.4
3
+ Version: 0.1.5
4
4
  Summary: Proteomics data loaders for AnnData format.
5
5
  Author: Max Frank
6
6
  License-Expression: MIT
@@ -46,9 +46,19 @@ Dynamic: license-file
46
46
  ![PyPI - Version](https://img.shields.io/pypi/v/protdata)
47
47
  [![docs online](https://img.shields.io/badge/docs-online-blue)](https://protdata-czbiohub.vercel.app/)
48
48
 
49
+ <a href="https://protdata.sf.czbiohub.org">
50
+ <img
51
+ src="https://raw.githubusercontent.com/czbiohub-sf/protdata/refs/heads/main/docs/source/_static/img/protdata_schema.svg"
52
+ class="dark-light"
53
+ align="right"
54
+ width="150"
55
+ alt="image"
56
+ />
57
+ </a>
58
+
49
59
  Proteomics data loaders for the [AnnData](https://anndata.readthedocs.io/) format.
50
60
 
51
- This package provides loader functions to import proteomics data (e.g., MaxQuant) into the AnnData structure for downstream analysis and integration with single-cell and multi-omics workflows.
61
+ This package provides loader functions to import proteomics data (e.g., MaxQuant) into the AnnData structure for downstream analysis and easy integration with single-cell and multi-omics workflows.
52
62
 
53
63
  ## Features
54
64
 
@@ -124,7 +134,7 @@ To get in touch please use the [GihHub issues](https://github.com/czbiohub-sf/pr
124
134
 
125
135
  ## Contributing
126
136
 
127
- If you want to contribute to `protdata`, please read the [Contribution Guide](https://prodata.sf.czbiohub.org/contributing)
137
+ If you want to contribute to `protdata`, please read the [Contribution Guide](https://protdata.sf.czbiohub.org/contributing.html)
128
138
 
129
139
  ## Changelog
130
- See [Release Notes](https://prodata.sf.czbiohub.org/release_notes)
140
+ See [Release Notes](https://protdata.sf.czbiohub.org/release_notes.html)
@@ -4,9 +4,19 @@
4
4
  ![PyPI - Version](https://img.shields.io/pypi/v/protdata)
5
5
  [![docs online](https://img.shields.io/badge/docs-online-blue)](https://protdata-czbiohub.vercel.app/)
6
6
 
7
+ <a href="https://protdata.sf.czbiohub.org">
8
+ <img
9
+ src="https://raw.githubusercontent.com/czbiohub-sf/protdata/refs/heads/main/docs/source/_static/img/protdata_schema.svg"
10
+ class="dark-light"
11
+ align="right"
12
+ width="150"
13
+ alt="image"
14
+ />
15
+ </a>
16
+
7
17
  Proteomics data loaders for the [AnnData](https://anndata.readthedocs.io/) format.
8
18
 
9
- This package provides loader functions to import proteomics data (e.g., MaxQuant) into the AnnData structure for downstream analysis and integration with single-cell and multi-omics workflows.
19
+ This package provides loader functions to import proteomics data (e.g., MaxQuant) into the AnnData structure for downstream analysis and easy integration with single-cell and multi-omics workflows.
10
20
 
11
21
  ## Features
12
22
 
@@ -82,7 +92,7 @@ To get in touch please use the [GihHub issues](https://github.com/czbiohub-sf/pr
82
92
 
83
93
  ## Contributing
84
94
 
85
- If you want to contribute to `protdata`, please read the [Contribution Guide](https://prodata.sf.czbiohub.org/contributing)
95
+ If you want to contribute to `protdata`, please read the [Contribution Guide](https://protdata.sf.czbiohub.org/contributing.html)
86
96
 
87
97
  ## Changelog
88
- See [Release Notes](https://prodata.sf.czbiohub.org/release_notes)
98
+ See [Release Notes](https://protdata.sf.czbiohub.org/release_notes.html)
@@ -62,12 +62,14 @@ issues_github_path = "czbiohub-sf/protdata"
62
62
 
63
63
  nitpicky = True
64
64
  nitpick_ignore = []
65
+ # scanpydoc is deprecated in favour of scverse-doc; silence the notice until we migrate
66
+ suppress_warnings = ["scanpydoc.deprecated"]
65
67
 
66
68
  intersphinx_mapping = dict(
67
69
  python=("https://docs.python.org/3", None),
68
70
  numpy=("https://numpy.org/doc/stable", None),
69
71
  pandas=("https://pandas.pydata.org/pandas-docs/stable", None),
70
- anndata=("https://anndata.readthedocs.io/en/stable/", None),
72
+ anndata=("https://anndata.scverse.org/en/stable/", None),
71
73
  scipy=("https://docs.scipy.org/doc/scipy", None),
72
74
  h5py=("https://docs.h5py.org/en/latest", None),
73
75
  )
@@ -0,0 +1,13 @@
1
+ from .diann_loader import read_diann
2
+ from .fragpipe_loader import read_fragpipe
3
+ from .maxquant_loader import read_maxquant
4
+ from .mztab_loader import read_mztab
5
+ from .spectronaut_loader import read_spectronaut
6
+
7
+ __all__ = [
8
+ "read_maxquant",
9
+ "read_diann",
10
+ "read_fragpipe",
11
+ "read_mztab",
12
+ "read_spectronaut",
13
+ ]
@@ -51,17 +51,26 @@ def read_maxquant(
51
51
  - ``var``: protein metadata (indexed by protein IDs)
52
52
  - ``obs``: sample metadata (indexed by sample names)
53
53
  - ``layers``: additional intensity matrices if multiple intensity column prefixes are provided
54
+
55
+ Notes
56
+ -----
57
+ - The first intensity column prefix is used for the main matrix (X), others are stored as layers if present.
58
+ - Forward slashes (``/``) are not allowed in hdf5 keys, so they are replaced with underscores (``_``).
54
59
  """
55
60
  if isinstance(intensity_column_prefixes, str):
56
61
  intensity_column_prefixes = [intensity_column_prefixes]
57
62
 
58
- main_intensity_column = intensity_column_prefixes[0]
59
63
  if isinstance(file, pd.DataFrame):
60
64
  df = file
61
65
  else:
62
66
  df = pd.read_csv(file, sep=sep, low_memory=False)
63
67
 
68
+ # Forward slashes are not allowed in hdf5 keys
69
+ df.columns = df.columns.str.replace("/", "_")
70
+ intensity_column_prefixes = [x.replace("/", "_") for x in intensity_column_prefixes]
71
+
64
72
  # Find intensity columns
73
+ main_intensity_column = intensity_column_prefixes[0]
65
74
  intensity_cols = [
66
75
  col for col in df.columns if col.startswith(main_intensity_column)
67
76
  ]
@@ -84,11 +93,11 @@ def read_maxquant(
84
93
  if col in [prefix + sample_name for sample_name in sample_names]
85
94
  ]
86
95
  if len(prefix_cols) == len(sample_names):
87
- # Cannot have '/' in the key (hdf5 interprets as a group for serialization)
88
- layers[prefix.strip().replace("/", "_")] = (
89
- df[prefix_cols].to_numpy(dtype=np.float32).T
90
- )
96
+ layers[prefix.strip()] = df[prefix_cols].to_numpy(dtype=np.float32).T
91
97
  else:
98
+ print(prefix_cols)
99
+ print(len(prefix_cols))
100
+ print(len(sample_names))
92
101
  warnings.warn(
93
102
  f"Number of columns for prefix '{prefix}' does not match number of samples."
94
103
  )
@@ -0,0 +1,167 @@
1
+ import warnings
2
+ from typing import Any, Dict, List, Optional, Tuple
3
+
4
+ import anndata as ad
5
+ import pandas as pd
6
+
7
+
8
+ def read_spectronaut(
9
+ file: str,
10
+ intensity_columns: Optional[List[str] | str] = ["PG.Quantity"],
11
+ index_column: Optional[str] = "PG.ProteinGroups",
12
+ sample_column: Optional[str] = "R.FileName",
13
+ sep="\t",
14
+ ):
15
+ """
16
+ Load Spectronaut results into an AnnData object.
17
+
18
+ Parameters
19
+ ----------
20
+ file
21
+ Path to the Spectronaut results file.
22
+ intensity_columns
23
+ Name of the intensity column.
24
+ index_column
25
+ Name of the column to use as protein index.
26
+ sample_column
27
+ Name of the column to use as sample index.
28
+ sep
29
+ File separator.
30
+
31
+ Returns
32
+ -------
33
+ :class:`anndata.AnnData` object with:
34
+
35
+ - ``X``: intensity matrix (samples x proteins)
36
+ - ``var``: protein metadata (indexed by protein group IDs)
37
+ - ``obs``: sample metadata (indexed by sample names)
38
+ """
39
+
40
+ if isinstance(intensity_columns, str):
41
+ intensity_columns = [intensity_columns]
42
+ # Check that intensity columns are all at the same level as the index column
43
+ index_level = index_column.split(".")[0]
44
+ intensity_levels = [col.split(".")[0] for col in intensity_columns]
45
+ if not all(level == index_level for level in intensity_levels):
46
+ raise ValueError(
47
+ f"Intensity columns {intensity_columns} are not all at the same level as the index column {index_column}"
48
+ )
49
+
50
+ sample_levels = ("E", "R")
51
+ possible_levels = ("PG", "PEP", "EG", "FG", "F")
52
+ df = _read_csv_auto_decimal(file, sep=sep)
53
+
54
+ sample_level = sample_column.split(".")[0]
55
+ sample_idx = sample_levels.index(sample_level)
56
+ var_levels = sample_levels[: sample_idx + 1]
57
+ var_levels = tuple(f"{level}." for level in var_levels)
58
+ var_cols = [
59
+ col
60
+ for col in df.columns
61
+ if col.startswith(var_levels) and df[col].notna().sum() > 0
62
+ ]
63
+
64
+ dfp = df.pivot_table(
65
+ index=index_column,
66
+ columns=var_cols,
67
+ values=intensity_columns,
68
+ aggfunc="mean",
69
+ observed=True,
70
+ dropna=True,
71
+ ).T
72
+ dfpx = dfp.loc[intensity_columns[0]]
73
+
74
+ # Report if intensities were not unique for pivoting and aggregated. This might not be what the user wants.
75
+ for intensity_column in intensity_columns:
76
+ if df[intensity_column].nunique() > dfpx.size:
77
+ warnings.warn(
78
+ f"{intensity_column} is not unique within {index_column} and {sample_column}, pivoting with mean aggregation. Make sure this is what you want!"
79
+ )
80
+
81
+ # Construct obs
82
+ obs = dfpx.index.to_frame().set_index(sample_column)
83
+
84
+ # Construct var
85
+ # Only columns that are coarser than the value column make sense to keep in var
86
+ var_levels = tuple(
87
+ f"{level}."
88
+ for level in possible_levels[: possible_levels.index(index_level) + 1]
89
+ )
90
+ varg = df.loc[:, df.columns.str.startswith(var_levels)].groupby(
91
+ index_column, observed=True, dropna=False
92
+ )
93
+ uniquecols = varg.nunique().eq(1).all(axis=0)
94
+ uniquecols = uniquecols[uniquecols].index
95
+ var = varg[uniquecols].first().reindex(dfpx.columns)
96
+
97
+ # Get the layers
98
+ layers = {}
99
+ for intensity_column in intensity_columns[1:]:
100
+ layers[intensity_column] = dfp.loc[intensity_column]
101
+
102
+ # Create AnnData
103
+ uns = {
104
+ "RawInfo": {
105
+ "Search_Engine": "Spectronaut",
106
+ },
107
+ }
108
+ adata = ad.AnnData(X=dfpx.to_numpy(), obs=obs, var=var, uns=uns, layers=layers)
109
+ return adata
110
+
111
+
112
+ def _read_csv_auto_decimal(
113
+ path: str,
114
+ *,
115
+ sample_rows: int = 2000,
116
+ trials: List[Dict[str, Optional[str]]] = [
117
+ {
118
+ "decimal": ".",
119
+ },
120
+ {
121
+ "decimal": ",",
122
+ },
123
+ ],
124
+ **kwargs: Any,
125
+ ) -> pd.DataFrame:
126
+ """
127
+ Read a CSV by trying multiple (decimal, thousands) configurations on a sample
128
+ using a fixed `sep` (no separator inference). Picks the best config and
129
+ re-reads the full file once.
130
+
131
+ Parameters
132
+ ----------
133
+ path : str
134
+ CSV file path.
135
+ sample_rows : int
136
+ Number of rows to sample for scoring.
137
+ trials : list of dict
138
+ List of {'decimal': <'.' or ','>, 'thousands': <'.' or ',' or None>} to try.
139
+ **kwargs : Any
140
+ Passed through to `pd.read_csv`.
141
+
142
+ Returns
143
+ -------
144
+ DataFrame
145
+ """
146
+
147
+ scores: List[Tuple[int, Dict[str, Optional[str]]]] = []
148
+ base_kwargs = dict(kwargs)
149
+
150
+ for cfg in trials:
151
+ try:
152
+ samp = pd.read_csv(path, nrows=sample_rows, **base_kwargs, **cfg)
153
+ num = samp.select_dtypes(include="number")
154
+ # Score = (non-NaN numeric cells)*10 + (# numeric cols)
155
+ frac_valid = 1.0 - num.isna().to_numpy().mean()
156
+ score = frac_valid * 10000 + num.shape[1] # or any large scale
157
+
158
+ scores.append((score, cfg))
159
+ except Exception:
160
+ scores.append((-1, cfg))
161
+
162
+ best_score, best_cfg = max(scores, key=lambda x: x[0])
163
+ if best_score < 0:
164
+ # All trials failed; fall back to single read with caller's sep and defaults
165
+ return pd.read_csv(path, **base_kwargs)
166
+
167
+ return pd.read_csv(path, **base_kwargs, **best_cfg)
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: protdata
3
- Version: 0.1.4
3
+ Version: 0.1.5
4
4
  Summary: Proteomics data loaders for AnnData format.
5
5
  Author: Max Frank
6
6
  License-Expression: MIT
@@ -46,9 +46,19 @@ Dynamic: license-file
46
46
  ![PyPI - Version](https://img.shields.io/pypi/v/protdata)
47
47
  [![docs online](https://img.shields.io/badge/docs-online-blue)](https://protdata-czbiohub.vercel.app/)
48
48
 
49
+ <a href="https://protdata.sf.czbiohub.org">
50
+ <img
51
+ src="https://raw.githubusercontent.com/czbiohub-sf/protdata/refs/heads/main/docs/source/_static/img/protdata_schema.svg"
52
+ class="dark-light"
53
+ align="right"
54
+ width="150"
55
+ alt="image"
56
+ />
57
+ </a>
58
+
49
59
  Proteomics data loaders for the [AnnData](https://anndata.readthedocs.io/) format.
50
60
 
51
- This package provides loader functions to import proteomics data (e.g., MaxQuant) into the AnnData structure for downstream analysis and integration with single-cell and multi-omics workflows.
61
+ This package provides loader functions to import proteomics data (e.g., MaxQuant) into the AnnData structure for downstream analysis and easy integration with single-cell and multi-omics workflows.
52
62
 
53
63
  ## Features
54
64
 
@@ -124,7 +134,7 @@ To get in touch please use the [GihHub issues](https://github.com/czbiohub-sf/pr
124
134
 
125
135
  ## Contributing
126
136
 
127
- If you want to contribute to `protdata`, please read the [Contribution Guide](https://prodata.sf.czbiohub.org/contributing)
137
+ If you want to contribute to `protdata`, please read the [Contribution Guide](https://protdata.sf.czbiohub.org/contributing.html)
128
138
 
129
139
  ## Changelog
130
- See [Release Notes](https://prodata.sf.czbiohub.org/release_notes)
140
+ See [Release Notes](https://protdata.sf.czbiohub.org/release_notes.html)
@@ -14,6 +14,8 @@ protdata/io/diann_loader.py
14
14
  protdata/io/fragpipe_loader.py
15
15
  protdata/io/maxquant_loader.py
16
16
  protdata/io/mztab_loader.py
17
+ protdata/io/spectronaut_loader.py
17
18
  protdata/io/utils.py
18
19
  tests/test_loaders.py
19
- tests/test_maxquant_loader.py
20
+ tests/test_maxquant_loader.py
21
+ tests/test_spectronaut_loader.py
@@ -1,6 +1,6 @@
1
1
  [project]
2
2
  name = "protdata"
3
- version = "0.1.4"
3
+ version = "0.1.5"
4
4
  description = "Proteomics data loaders for AnnData format."
5
5
  requires-python = ">=3.9, <4.0"
6
6
  authors = [
@@ -7,6 +7,7 @@ from protdata.io.diann_loader import read_diann
7
7
  from protdata.io.fragpipe_loader import read_fragpipe
8
8
  from protdata.io.maxquant_loader import read_maxquant
9
9
  from protdata.io.mztab_loader import read_mztab
10
+ from protdata.io.spectronaut_loader import read_spectronaut
10
11
 
11
12
  data_dir = os.path.abspath(os.path.join(os.path.dirname(__file__), "../data"))
12
13
 
@@ -18,6 +19,7 @@ data_dir = os.path.abspath(os.path.join(os.path.dirname(__file__), "../data"))
18
19
  ("combined_protein.tsv", read_fragpipe),
19
20
  ("SILAC_SQ.mzTab", read_mztab),
20
21
  ("report.pg_matrix.tsv", read_diann),
22
+ ("spectronaut_normal.tsv", read_spectronaut),
21
23
  ],
22
24
  )
23
25
  def test_loader(filename, loader, tmp_path):
@@ -0,0 +1,97 @@
1
+ import os
2
+ import warnings
3
+
4
+ import numpy as np
5
+ import pandas as pd
6
+ import pytest
7
+
8
+ from protdata.io.spectronaut_loader import read_spectronaut
9
+
10
+ # Absolute path provided by the user for local validation
11
+ LOCAL_DATA_PATH = os.path.abspath(
12
+ os.path.join(os.path.dirname(__file__), "../data/spectronaut_normal.tsv")
13
+ )
14
+
15
+
16
+ def test_spectronaut_pivot_warning_and_shape():
17
+ with pytest.warns(UserWarning, match="pivoting with mean aggregation"):
18
+ adata = read_spectronaut(
19
+ LOCAL_DATA_PATH,
20
+ index_column="EG.Workflow",
21
+ intensity_columns=["EG.Cscore", "EG.Qvalue"],
22
+ )
23
+ assert adata.shape == (9, 1)
24
+
25
+
26
+ def test_spectronaut_no_warning_and_shape():
27
+ with warnings.catch_warnings(record=True) as caught:
28
+ warnings.simplefilter("always")
29
+ adata = read_spectronaut(
30
+ LOCAL_DATA_PATH,
31
+ index_column="EG.ModifiedSequence",
32
+ intensity_columns=["EG.Cscore", "EG.Qvalue"],
33
+ )
34
+ assert len(caught) == 0
35
+ assert adata.shape == (9, 14)
36
+
37
+
38
+ def _count_nonnull_grouped_values(
39
+ df: pd.DataFrame, index_col: str, sample_col: str, intensity_col: str
40
+ ) -> int:
41
+ grouped = (
42
+ df[[index_col, sample_col, intensity_col]]
43
+ .groupby([index_col, sample_col], observed=True, dropna=False)[intensity_col]
44
+ .apply(lambda s: pd.notna(s).any())
45
+ )
46
+ return int(grouped.sum())
47
+
48
+
49
+ def _count_nonnull_in_matrix(matrix: np.ndarray) -> int:
50
+ return int(np.count_nonzero(~np.isnan(matrix)))
51
+
52
+
53
+ def test_spectronaut_pivot_nonnull_counts_match_local_data():
54
+ assert os.path.isfile(LOCAL_DATA_PATH)
55
+ # Read raw for ground truth counting (file uses comma decimals)
56
+ df = pd.read_csv(LOCAL_DATA_PATH, sep="\t", decimal=",")
57
+ adata = read_spectronaut(LOCAL_DATA_PATH)
58
+
59
+ grouped_nonnull = _count_nonnull_grouped_values(
60
+ df,
61
+ index_col="PG.ProteinGroups",
62
+ sample_col="R.FileName",
63
+ intensity_col="PG.Quantity",
64
+ )
65
+ matrix_nonnull = _count_nonnull_in_matrix(adata.X)
66
+
67
+ assert matrix_nonnull == grouped_nonnull
68
+
69
+
70
+ def test_spectronaut_layers_shape_and_presence():
71
+ assert os.path.isfile(LOCAL_DATA_PATH)
72
+ adata = read_spectronaut(
73
+ LOCAL_DATA_PATH,
74
+ intensity_columns=["PG.Quantity", "PG.Cscore"],
75
+ )
76
+ assert "PG.Cscore" in adata.layers
77
+ assert adata.layers["PG.Cscore"].shape == adata.X.shape
78
+
79
+
80
+ def test_spectronaut_uns_and_obs_metadata():
81
+ assert os.path.isfile(LOCAL_DATA_PATH)
82
+ adata = read_spectronaut(LOCAL_DATA_PATH)
83
+ assert adata.uns["RawInfo"]["Search_Engine"] == "Spectronaut"
84
+ # obs index is the sample identifier column
85
+ assert adata.obs.index.name == "R.FileName"
86
+ # expect at least one other sample-level column to be present
87
+ assert "R.Replicate" in adata.obs.columns or "R.Condition" in adata.obs.columns
88
+
89
+
90
+ def test_spectronaut_level_mismatch_raises():
91
+ assert os.path.isfile(LOCAL_DATA_PATH)
92
+ with pytest.raises(ValueError):
93
+ read_spectronaut(
94
+ LOCAL_DATA_PATH,
95
+ index_column="EG.Workflow",
96
+ intensity_columns=["PG.Quantity"],
97
+ )
@@ -1,6 +0,0 @@
1
- from .maxquant_loader import read_maxquant
2
- from .diann_loader import read_diann
3
- from .fragpipe_loader import read_fragpipe
4
- from .mztab_loader import read_mztab
5
-
6
- __all__ = ["read_maxquant", "read_diann", "read_fragpipe", "read_mztab"]
File without changes
File without changes
File without changes