sclab 0.4.0__tar.gz → 0.4.2__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.
Files changed (102) hide show
  1. {sclab-0.4.0 → sclab-0.4.2}/PKG-INFO +3 -3
  2. {sclab-0.4.0 → sclab-0.4.2}/pyproject.toml +11 -10
  3. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/__init__.py +1 -1
  4. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_transfer_metadata.py +15 -2
  5. sclab-0.4.2/src/sclab/tools/__init__.py +7 -0
  6. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/_density_dynamics.py +1 -1
  7. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/utils/density_nd.py +1 -1
  8. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/utils/periodic_genes.py +9 -2
  9. sclab-0.4.2/src/sclab/tools/differential_expression/__init__.py +11 -0
  10. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/differential_expression/_pseudobulk_edger.py +20 -5
  11. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/differential_expression/_pseudobulk_limma.py +20 -5
  12. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/doublet_detection/_scdblfinder.py +19 -5
  13. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/labeling/sctype.py +12 -0
  14. sclab-0.4.2/src/sclab/tools/utils/__init__.py +7 -0
  15. sclab-0.4.2/src/sclab/tools/utils/_call_differential_expression.py +88 -0
  16. sclab-0.4.0/src/sclab/tools/differential_expression/__init__.py +0 -7
  17. sclab-0.4.0/src/sclab/tools/imputation/__init__.py +0 -0
  18. sclab-0.4.0/src/sclab/tools/utils/__init__.py +0 -5
  19. {sclab-0.4.0 → sclab-0.4.2}/LICENSE +0 -0
  20. {sclab-0.4.0 → sclab-0.4.2}/README.md +0 -0
  21. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/_io.py +0 -0
  22. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/_methods_registry.py +0 -0
  23. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/_sclab.py +0 -0
  24. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/__init__.py +0 -0
  25. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/_dataset.py +0 -0
  26. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/_exceptions.py +0 -0
  27. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/plotter/__init__.py +0 -0
  28. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/plotter/_controls.py +0 -0
  29. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/plotter/_plotter.py +0 -0
  30. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/plotter/_utils.py +0 -0
  31. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/processor/__init__.py +0 -0
  32. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/processor/_processor.py +0 -0
  33. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/processor/_results_panel.py +0 -0
  34. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/processor/step/__init__.py +0 -0
  35. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/processor/step/_basic_processor_step.py +0 -0
  36. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/dataset/processor/step/_processor_step_base.py +0 -0
  37. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/event/__init__.py +0 -0
  38. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/event/_broker.py +0 -0
  39. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/event/_client.py +0 -0
  40. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/event/_utils.py +0 -0
  41. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/__init__.py +0 -0
  42. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/__init__.py +0 -0
  43. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_cluster.py +0 -0
  44. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_differential_expression.py +0 -0
  45. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_doublet_detection.py +0 -0
  46. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_gene_expression.py +0 -0
  47. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_integration.py +0 -0
  48. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_neighbors.py +0 -0
  49. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_pca.py +0 -0
  50. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_preprocess.py +0 -0
  51. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_qc.py +0 -0
  52. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/examples/processor_steps/_umap.py +0 -0
  53. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/gui/__init__.py +0 -0
  54. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/gui/components/__init__.py +0 -0
  55. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/gui/components/_guided_pseudotime.py +0 -0
  56. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/gui/components/_transfer_metadata.py +0 -0
  57. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/methods/__init__.py +0 -0
  58. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/__init__.py +0 -0
  59. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_cca.py +0 -0
  60. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_cca_integrate.py +0 -0
  61. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_filter_obs.py +0 -0
  62. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_harmony.py +0 -0
  63. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_harmony_integrate.py +0 -0
  64. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_mcca.py +0 -0
  65. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_mcca2.py +0 -0
  66. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_normalize_weighted.py +0 -0
  67. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_pca.py +0 -0
  68. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_preprocess.py +0 -0
  69. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_qc.py +0 -0
  70. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_rpca.py +0 -0
  71. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_subset.py +0 -0
  72. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_transform.py +0 -0
  73. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/preprocess/_utils.py +0 -0
  74. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/__init__.py +0 -0
  75. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/_compat.py +0 -0
  76. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/_settings.py +0 -0
  77. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/logging.py +0 -0
  78. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/plotting/__init__.py +0 -0
  79. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/plotting/_rcmod.py +0 -0
  80. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/plotting/palettes.py +0 -0
  81. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/scanpy/readwrite.py +0 -0
  82. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/__init__.py +0 -0
  83. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/_expression_dynamics.py +0 -0
  84. {sclab-0.4.0/src/sclab/tools → sclab-0.4.2/src/sclab/tools/cellflow/_pseudotime}/__init__.py +0 -0
  85. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/_pseudotime/_pseudotime.py +0 -0
  86. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/_pseudotime/_timeseries.py +0 -0
  87. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/pseudotime_tools/__init__.py +0 -0
  88. {sclab-0.4.0/src/sclab/tools/cellflow/_pseudotime → sclab-0.4.2/src/sclab/tools/cellflow/utils}/__init__.py +0 -0
  89. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/utils/interpolate.py +0 -0
  90. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/utils/smoothen.py +0 -0
  91. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/cellflow/utils/times.py +0 -0
  92. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/differential_expression/_pseudobulk_helpers.py +0 -0
  93. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/doublet_detection/__init__.py +0 -0
  94. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/doublet_detection/_doubletdetection.py +0 -0
  95. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/doublet_detection/_scrublet.py +0 -0
  96. {sclab-0.4.0/src/sclab/tools/cellflow/utils → sclab-0.4.2/src/sclab/tools/embedding}/__init__.py +0 -0
  97. {sclab-0.4.0/src/sclab/tools/embedding → sclab-0.4.2/src/sclab/tools/imputation}/__init__.py +0 -0
  98. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/imputation/_alra.py +0 -0
  99. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/labeling/__init__.py +0 -0
  100. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/tools/utils/_aggregate_and_filter.py +0 -0
  101. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/utils/__init__.py +0 -0
  102. {sclab-0.4.0 → sclab-0.4.2}/src/sclab/utils/_write_excel.py +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: sclab
3
- Version: 0.4.0
3
+ Version: 0.4.2
4
4
  Summary: sclab
5
5
  Author-email: Argenis Arriojas <ArriojasMaldonado001@umb.edu>
6
6
  Requires-Python: >=3.10,<=3.13
@@ -21,8 +21,8 @@ Requires-Dist: matplotlib
21
21
  Requires-Dist: numpy
22
22
  Requires-Dist: pandas
23
23
  Requires-Dist: plotly<6.0
24
+ Requires-Dist: pydeseq2>=0.5.2
24
25
  Requires-Dist: requests
25
- Requires-Dist: ripser>=0.6.12
26
26
  Requires-Dist: scikit-learn
27
27
  Requires-Dist: scipy
28
28
  Requires-Dist: svgpathtools
@@ -34,7 +34,7 @@ Requires-Dist: scanpy[leiden, skmisc]>=1.10 ; extra == "scanpy"
34
34
  Requires-Dist: pytest>=8.3.4 ; extra == "test"
35
35
  Project-URL: Bug Tracker, https://github.com/umbibio/sclab/issues
36
36
  Project-URL: Changelog, https://github.com/umbibio/sclab/blob/main/CHANGELOG.md
37
- Project-URL: Documentation, https://github.com/umbibio/sclab/docs
37
+ Project-URL: Documentation, https://sclab.readthedocs.io
38
38
  Project-URL: Homepage, https://github.com/umbibio/sclab
39
39
  Project-URL: Repository, https://github.com/umbibio/sclab.git
40
40
  Provides-Extra: jupyter
@@ -1,6 +1,6 @@
1
1
  [project]
2
2
  name = "sclab"
3
- version = "0.4.0"
3
+ version = "0.4.2"
4
4
  description = "sclab"
5
5
  readme = "README.md"
6
6
  authors = [
@@ -25,8 +25,8 @@ dependencies = [
25
25
  "numpy",
26
26
  "pandas",
27
27
  "plotly<6.0",
28
+ "pydeseq2>=0.5.2",
28
29
  "requests",
29
- "ripser>=0.6.12",
30
30
  "scikit-learn",
31
31
  "scipy",
32
32
  "svgpathtools",
@@ -35,7 +35,7 @@ dependencies = [
35
35
 
36
36
  [project.urls]
37
37
  "Repository" = "https://github.com/umbibio/sclab.git"
38
- "Documentation" = "https://github.com/umbibio/sclab/docs"
38
+ "Documentation" = "https://sclab.readthedocs.io"
39
39
  "Homepage" = "https://github.com/umbibio/sclab"
40
40
  "Bug Tracker" = "https://github.com/umbibio/sclab/issues"
41
41
  "Changelog" = "https://github.com/umbibio/sclab/blob/main/CHANGELOG.md"
@@ -54,23 +54,24 @@ build-backend = "flit_core.buildapi"
54
54
  name = "sclab"
55
55
  source = "src/sclab"
56
56
 
57
- [tool.uv.sources]
58
- sclab-tools = { path = "../sclab-tools" }
59
-
60
57
  [dependency-groups]
58
+ docs = [
59
+ "mkdocs>=1.6.1",
60
+ "mkdocs-material>=9.6.3",
61
+ "mkdocs-jupyter>=0.25.1",
62
+ "mkdocstrings[python]>=1.0.3",
63
+ ]
61
64
  dev = [
62
65
  "bump-my-version>=0.31.1",
63
66
  "pre-commit>=4.1.0",
64
67
  "ruff>=0.9.4",
65
68
  "nox>=2024.1.29",
66
69
  "pytest>=8.3.4",
67
- "mkdocs>=1.6.1",
68
- "mkdocs-material>=9.6.3",
69
- "mkdocs-jupyter>=0.25.1",
70
70
  "scanpy[harmony,leiden,scrublet,skmisc]>=1.10.4",
71
71
  "jupyterlab>=4.3.6",
72
- "sclab-tools",
73
72
  "scrublet>=0.2.3",
74
73
  "doubletdetection>=4.3.0.post1",
75
74
  "sclab[jupyter,r,scanpy]",
75
+ "ripser>=0.6.12",
76
+ { include-group = "docs" },
76
77
  ]
@@ -6,4 +6,4 @@ __all__ = [
6
6
  "SCLabDashboard",
7
7
  ]
8
8
 
9
- __version__ = "0.4.0"
9
+ __version__ = "0.4.2"
@@ -238,12 +238,25 @@ def _assign_categorical(values: pd.Series, weights: NDArray):
238
238
  def _assign_numerical(values: pd.Series, weights: NDArray):
239
239
  # weighted mean and standard error
240
240
  sum_w: float = weights.sum()
241
- sum2_w: float = weights.sum() ** 2
241
+
242
+ # If weights sum to 0, mean and error are undefined
243
+ if sum_w == 0:
244
+ return np.nan, np.nan
245
+
246
+ sum2_w: float = sum_w ** 2
242
247
  sum_w2: float = (weights**2).sum()
243
248
  n_eff: float = sum2_w / sum_w2
244
249
 
245
250
  mean_x: float = (values * weights).sum() / sum_w
246
- var_x: float = ((values - mean_x) ** 2 * weights).sum() * sum_w / (sum2_w - sum_w2)
251
+
252
+ var_denom = sum2_w - sum_w2
253
+
254
+ # If effective sample size is 1, variance is undefined (denominator is 0)
255
+ # We use <= 0 to handle potential floating point inaccuracies
256
+ if var_denom <= 0:
257
+ return mean_x, np.nan
258
+
259
+ var_x: float = ((values - mean_x) ** 2 * weights).sum() * sum_w / var_denom
247
260
  err_x: float = np.sqrt(var_x / n_eff)
248
261
 
249
262
  return mean_x, err_x
@@ -0,0 +1,7 @@
1
+ from . import cellflow, differential_expression, labeling
2
+
3
+ __all__ = [
4
+ "cellflow",
5
+ "differential_expression",
6
+ "labeling",
7
+ ]
@@ -146,7 +146,7 @@ def density(
146
146
  )
147
147
 
148
148
  if plot_density | plot_density_fit | plot_density_fit_derivative | plot_histogram:
149
- from ..utils.density_nd import density_result_1d
149
+ from .utils.density_nd import density_result_1d
150
150
 
151
151
  density_result_1d(
152
152
  rslt,
@@ -184,7 +184,7 @@ def density_result_1d(
184
184
  # we recover the original time vector and corresponding density
185
185
  x = xx[histogram_nbins : 2 * histogram_nbins]
186
186
  d = dd[histogram_nbins : 2 * histogram_nbins] * 3 # correct the density
187
- ax.bar(x, d, width=1 / histogram_nbins, fill=False, linewidth=0.5)
187
+ ax.bar(x, d, width=(tmax - tmin) / histogram_nbins, fill=False, linewidth=0.5)
188
188
 
189
189
  x = np.linspace(tmin, tmax, 2**10 + 1)
190
190
  if plot_density_fit:
@@ -18,6 +18,8 @@ def periodic_genes(
18
18
  min_pct_power_below: float = 0.75,
19
19
  layer: str | None = None,
20
20
  ):
21
+ import scanpy as sc
22
+
21
23
  times = adata.obs[time_key].values.copy()
22
24
  if layer is None or layer == "X":
23
25
  X = adata.X
@@ -41,9 +43,10 @@ def periodic_genes(
41
43
  make_stats=False,
42
44
  make_dummies=False,
43
45
  )
46
+ sc.pp.normalize_total(aggregated, target_sum=1e4)
44
47
  log_cnts = np.log1p(aggregated.X)
45
48
  profiles = pd.DataFrame(log_cnts, index=labels, columns=aggregated.var_names)
46
- ps = power_spectrum_df(profiles)
49
+ ps = power_spectrum_df(profiles, window="boxcar", detrend="constant")
47
50
  pp = pct_power_below(ps, 1 / period)
48
51
 
49
52
  adata.varm["profile"] = profiles.T
@@ -69,7 +72,11 @@ def infer_dt_from_index(idx: pd.Index) -> float:
69
72
  return dt
70
73
 
71
74
 
72
- def power_spectrum_df(X: pd.DataFrame, window: str = "hann", detrend: str = "constant"):
75
+ def power_spectrum_df(
76
+ X: pd.DataFrame,
77
+ window: str = "boxcar",
78
+ detrend: str = "constant",
79
+ ):
73
80
  # X: rows=timepoints, columns=variables
74
81
  Xd = X - X.mean() # remove DC so percent computations are stable
75
82
  dt = infer_dt_from_index(X.index) if X.index.size > 1 else 1.0
@@ -0,0 +1,11 @@
1
+ from ..utils._call_differential_expression import call_differential_expression
2
+ from ._pseudobulk_edger import edger_is_available, pseudobulk_edger
3
+ from ._pseudobulk_limma import limma_is_available, pseudobulk_limma
4
+
5
+ __all__ = [
6
+ "call_differential_expression",
7
+ "pseudobulk_edger",
8
+ "pseudobulk_limma",
9
+ "edger_is_available",
10
+ "limma_is_available",
11
+ ]
@@ -1,15 +1,16 @@
1
+ from importlib.util import find_spec
2
+
1
3
  import pandas as pd
2
4
  from anndata import AnnData
3
5
 
4
6
  from ..utils import aggregate_and_filter
5
7
 
8
+ PYTHON_DEPENDENCIES = ["rpy2", "anndata2ri"]
9
+ R_DEPENDENCIES = ["edgeR", "MAST", "SingleCellExperiment"]
10
+
6
11
 
7
12
  def edger_is_available() -> bool:
8
- try:
9
- _try_imports()
10
- except ImportError:
11
- return False
12
- return True
13
+ return _python_deps_available() and _r_deps_available()
13
14
 
14
15
 
15
16
  def pseudobulk_edger(
@@ -315,3 +316,17 @@ def _try_imports():
315
316
  )
316
317
  print(message)
317
318
  raise ImportError(message)
319
+
320
+
321
+ def _python_deps_available() -> bool:
322
+ return all([find_spec(dep) is not None for dep in PYTHON_DEPENDENCIES])
323
+
324
+
325
+ def _r_deps_available() -> bool:
326
+ if not find_spec("rpy2"):
327
+ return False
328
+
329
+ from rpy2.robjects import r
330
+
331
+ installed_packages = list(r("installed.packages()"))
332
+ return all([dep in installed_packages for dep in R_DEPENDENCIES])
@@ -1,15 +1,16 @@
1
+ from importlib.util import find_spec
2
+
1
3
  import pandas as pd
2
4
  from anndata import AnnData
3
5
 
4
6
  from ..utils import aggregate_and_filter
5
7
 
8
+ PYTHON_DEPENDENCIES = ["rpy2", "anndata2ri"]
9
+ R_DEPENDENCIES = ["edgeR", "limma", "MAST", "SingleCellExperiment"]
10
+
6
11
 
7
12
  def limma_is_available() -> bool:
8
- try:
9
- _try_imports()
10
- except ImportError:
11
- return False
12
- return True
13
+ return _python_deps_available() and _r_deps_available()
13
14
 
14
15
 
15
16
  def pseudobulk_limma(
@@ -327,3 +328,17 @@ def _try_imports():
327
328
  )
328
329
  print(message)
329
330
  raise ImportError(message)
331
+
332
+
333
+ def _python_deps_available() -> bool:
334
+ return all([find_spec(dep) is not None for dep in PYTHON_DEPENDENCIES])
335
+
336
+
337
+ def _r_deps_available() -> bool:
338
+ if not find_spec("rpy2"):
339
+ return False
340
+
341
+ from rpy2.robjects import r
342
+
343
+ installed_packages = list(r("installed.packages()"))
344
+ return all([dep in installed_packages for dep in R_DEPENDENCIES])
@@ -1,16 +1,16 @@
1
+ from importlib.util import find_spec
1
2
  from typing import Literal
2
3
 
3
4
  import numpy as np
4
5
  import pandas as pd
5
6
  from anndata import AnnData
6
7
 
8
+ PYTHON_DEPENDENCIES = ["rpy2", "anndata2ri"]
9
+ R_DEPENDENCIES = ["scater", "scDblFinder", "BiocParallel", "SingleCellExperiment"]
10
+
7
11
 
8
12
  def scdblfinder_is_available() -> bool:
9
- try:
10
- _try_imports()
11
- except ImportError:
12
- return False
13
- return True
13
+ return _python_deps_available() and _r_deps_available()
14
14
 
15
15
 
16
16
  def scdblfinder(
@@ -184,3 +184,17 @@ def _try_imports():
184
184
  )
185
185
  print(message)
186
186
  raise ImportError(message)
187
+
188
+
189
+ def _python_deps_available() -> bool:
190
+ return all([find_spec(dep) is not None for dep in PYTHON_DEPENDENCIES])
191
+
192
+
193
+ def _r_deps_available() -> bool:
194
+ if not find_spec("rpy2"):
195
+ return False
196
+
197
+ from rpy2.robjects import r
198
+
199
+ installed_packages = list(r("installed.packages()"))
200
+ return all([dep in installed_packages for dep in R_DEPENDENCIES])
@@ -144,6 +144,18 @@ def classify_cells(
144
144
  distances in `adata.uns[neighbors_key]["distances"]`. Defaults to `None`.
145
145
  save_scores
146
146
  Whether to save the classification scores. Defaults to `False`
147
+
148
+ Returns
149
+ -------
150
+ None
151
+ Results are written in place to `adata.obs[key_added]` (category
152
+ dtype, `pd.NA` for calls below `threshold`) and
153
+ `adata.obs[key_added + "_noNA"]` (best-guess label regardless of
154
+ confidence). `key_added` defaults to `marker_class_key` when not
155
+ given. If `save_scores=True`, also writes
156
+ `adata.obs[key_added + "_score"]` (max per-cell confidence score) and
157
+ `adata.obsm[key_added + "_scores"]` (full class-by-cell score
158
+ matrix).
147
159
  """
148
160
  # cite("10.1038/s41467-022-28803-w", __package__)
149
161
 
@@ -0,0 +1,7 @@
1
+ from ._aggregate_and_filter import aggregate_and_filter
2
+ from ._call_differential_expression import call_differential_expression
3
+
4
+ __all__ = [
5
+ "aggregate_and_filter",
6
+ "call_differential_expression",
7
+ ]
@@ -0,0 +1,88 @@
1
+ import numpy as np
2
+ from pandas import DataFrame, Series
3
+ from scipy.stats import median_abs_deviation
4
+
5
+
6
+ def _robust_z(x: Series):
7
+ # DOI: 10.1080/01621459.1993.10476408
8
+ mad: float = median_abs_deviation(x)
9
+ Sr = 1.4826 * mad
10
+ Zr = (x - x.median()) / Sr
11
+
12
+ return Zr
13
+
14
+
15
+ def call_differential_expression(
16
+ table: DataFrame,
17
+ pvalue_col: str,
18
+ logfc_col: str,
19
+ pct_col_prefix: str = "pct_",
20
+ max_pval: float = 0.05,
21
+ min_robust_z_level: float = 2.5,
22
+ min_pct: float = 0.05,
23
+ contrast_key: str | None = None,
24
+ copy: bool = False,
25
+ ):
26
+ """Call differentially expressed genes from robust Z-scores of log-fold-change.
27
+
28
+ Flags each row in `table` as up- or down-regulated by combining a p-value
29
+ cutoff, a robust Z-score threshold on `logfc_col` (median-absolute-deviation
30
+ based; DOI: 10.1080/01621459.1993.10476408), and a minimum expression
31
+ percentage across any `pct_*`-prefixed column.
32
+
33
+ Parameters
34
+ ----------
35
+ table : DataFrame
36
+ Differential expression results, one row per gene/feature.
37
+ pvalue_col : str
38
+ Column in `table` with (adjusted) p-values.
39
+ logfc_col : str
40
+ Column in `table` with log-fold-changes.
41
+ pct_col_prefix : str, optional
42
+ Prefix of columns giving percent-expressed values; the max across all
43
+ matching columns is used as the expression filter. Defaults to "pct_".
44
+ max_pval : float, optional
45
+ Maximum p-value for a gene to be called DE. Defaults to 0.05.
46
+ min_robust_z_level : float, optional
47
+ Minimum absolute robust Z-score of `logfc_col` for a gene to be called
48
+ DE. Defaults to 2.5.
49
+ min_pct : float, optional
50
+ Minimum percent-expressed value for a gene to be called DE. Defaults
51
+ to 0.05.
52
+ contrast_key : str or None, optional
53
+ If provided, robust Z-scores are computed independently within each
54
+ `table.groupby(contrast_key)` group rather than across the whole
55
+ table. Defaults to None.
56
+ copy : bool, optional
57
+ If True, operate on and return a copy of `table` instead of modifying
58
+ it in place. Defaults to False.
59
+
60
+ Returns
61
+ -------
62
+ DataFrame or None
63
+ `table` with two new columns: `robust_Z` (the computed robust
64
+ Z-score) and `DE` (1 up-regulated, -1 down-regulated, 0 not
65
+ significant). Returned only if `copy=True`; otherwise `table` is
66
+ modified in place and `None` is returned.
67
+ """
68
+ if copy:
69
+ table = table.copy()
70
+
71
+ pct_expr_high = table.filter(regex=f"^{pct_col_prefix}").max(axis=1)
72
+
73
+ if contrast_key is not None:
74
+ table["robust_Z"] = (
75
+ table.groupby(contrast_key)[logfc_col].apply(_robust_z).droplevel(0)
76
+ )
77
+ else:
78
+ table["robust_Z"] = _robust_z(table[logfc_col])
79
+
80
+ mask = table[pvalue_col] < max_pval
81
+ mask &= table["robust_Z"].abs() > min_robust_z_level
82
+ mask &= pct_expr_high > min_pct
83
+
84
+ table["DE"] = 0
85
+ table.loc[mask, "DE"] = table["robust_Z"].map(np.sign).astype(int)
86
+
87
+ if copy:
88
+ return table
@@ -1,7 +0,0 @@
1
- from ._pseudobulk_edger import pseudobulk_edger
2
- from ._pseudobulk_limma import pseudobulk_limma
3
-
4
- __all__ = [
5
- "pseudobulk_edger",
6
- "pseudobulk_limma",
7
- ]
File without changes
@@ -1,5 +0,0 @@
1
- from ._aggregate_and_filter import aggregate_and_filter
2
-
3
- __all__ = [
4
- "aggregate_and_filter",
5
- ]
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes