Merge pull request #7 from lzj1769/develop

Develop

.github/workflows/python-package.yml

@@ -26,6 +26,7 @@ jobs:
         python-version: ${{ matrix.python-version }}
     - name: Install dependencies
       run: |
+        python -m pip install scikit-learn
         python -m pip install --upgrade pip
         python -m pip install flake8 pytest
         python -m pip install ./

docs/source/notebooks/compare_with_chromVAR.ipynb

@@ -5,7 +5,7 @@
    "id": "33fe3950-e22b-4a28-a6dc-693fc6592dc4",
    "metadata": {},
    "source": [
-    "# Comparison between pychromVAR and chromVAR: Part II"
+    "# Compare with chromVAR: Part II"
    ]
   },
   {
@@ -50,7 +50,7 @@
     {
      "data": {
       "text/plain": [
-       "'0.0.2'"
+       "'0.0.3'"
       ]
      },
      "execution_count": 2,
@@ -92,6 +92,16 @@
     "adata"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "0c5b5f7b-79d1-4e8d-bf09-b255c0d088dd",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "adata.write_h5ad(\"example_data.h5ad\")"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "2bd28801-44b0-4a30-8bb3-230d1d33feff",
@@ -1049,7 +1059,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.12"
+   "version": "3.9.15"
   }
  },
  "nbformat": 4,

pychromvar/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.0.3"
+__version__ = "0.0.4"
 __version_info__ = tuple([int(num) for num in __version__.split('.')])  # noqa: F401
 
 from .preprocessing import get_bg_peaks, add_gc_bias, add_peak_seq

pychromvar/compute_deviations.py

@@ -49,7 +49,7 @@ def compute_deviations(data: Union[AnnData, MuData], n_jobs=-1) -> AnnData:
     logging.info('computing observed motif deviations...')
     motif_match = adata.varm['motif_match'].transpose()
 
-    obs_dev = _compute_dev(
+    obs_dev = _compute_deviations(
         (motif_match, adata.X.transpose(), expectation.transpose())).transpose()
 
     # compute background deviations for bias-correction
@@ -67,7 +67,7 @@ def compute_deviations(data: Union[AnnData, MuData], n_jobs=-1) -> AnnData:
             bg_peak_idx = adata.varm['bg_peaks'][:, i]
             bg_motif_match = adata.varm['motif_match'][bg_peak_idx, :].transpose(
             )
-            bg_dev[i, :, :] = _compute_dev((bg_motif_match, adata.X.transpose(),
+            bg_dev[i, :, :] = _compute_deviations((bg_motif_match, adata.X.transpose(),
                                             expectation.transpose())).transpose()
 
     elif n_jobs > 1:
@@ -83,7 +83,7 @@ def compute_deviations(data: Union[AnnData, MuData], n_jobs=-1) -> AnnData:
 
         # run the function with multiple cpus
         with Pool(processes=n_jobs) as pool:
-            all_results = pool.map(_compute_dev, arguments_list)
+            all_results = pool.map(_compute_deviations, arguments_list)
 
         # parse the results
         for i in range(n_bg_peaks):
@@ -102,7 +102,7 @@ def compute_deviations(data: Union[AnnData, MuData], n_jobs=-1) -> AnnData:
     return dev
 
 
-def _compute_dev(arguments):
+def _compute_deviations(arguments):
     motif_match, count, expectation = arguments
 
     observed = np.dot(motif_match, count)
@@ -134,3 +134,4 @@ def compute_expectation(count: np.array) -> np.array:
     exp = np.dot(b, a)
 
     return exp
+

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "flit_core.buildapi"
 
 [project]
 name = "pychromvar"
-version = "0.0.3"
+version = "0.0.4"
 description = "A python package for chromVAR"
 authors = [
     {name = "Zhijian Li", email = "[email protected]"}
@@ -18,14 +18,14 @@ classifiers = [
     "Topic :: Scientific/Engineering :: Bio-Informatics",
     "Intended Audience :: Science/Research"
 ]
-requires-python = ">= 3.6"
+requires-python = ">= 3.8"
 dependencies = [
+    "scikit-learn",
     "anndata",
     "numpy",
     "scipy",
     "mudata",
     "scanpy",
-    "scikit-learn",
     "muon",
     "biopython",
     "MOODS-python",

tests/test_compute_deviations.py

@@ -1,6 +1,5 @@
 import numpy as np
-
-from pychromvar.compute_deviations import compute_expectation
+from pychromvar.compute_deviations import compute_expectation, _compute_deviations
 
 def test_compute_expectation():
     count = np.array([[1, 0, 1, ], [0, 1, 1]], dtype=np.float32)
@@ -11,3 +10,17 @@ def test_compute_expectation():
 
     # check the output
     assert np.array_equal(exp, np.array([[0.5, 0.5, 1], [0.5, 0.5, 1]]))
+
+def test_compute_deviations():
+    count = np.array([[1, 0, 1, ], [0, 1, 1]], dtype=np.float32)
+    exp = compute_expectation(count)
+
+    motif_match = np.array([[1, 1], [0, 1], [1, 0]], dtype=np.int8)
+
+    dev = _compute_deviations((motif_match, count, exp)).transpose()
+
+    # make sure output has same dimensionas
+    assert dev.shape == (count.shape[1], motif_match.shape[0])
+
+    # check the output
+    assert np.array_equal(dev, np.array([[0.0, -1, 1], [0, 1, -1], [0, 0, 0]]))

tests/test_preprocessing.py

@@ -0,0 +1,11 @@
+import numpy as np
+from anndata import AnnData
+
+from pychromvar.preprocessing import add_gc_bias
+
+def test_gc_bias():
+    data = AnnData(np.array([[1, 2,], [3, 4]]), dtype=np.float32)
+    data.uns["peak_seq"] = ["AAAAAAAA", "CCCCGGGG"]
+    add_gc_bias(data=data)
+
+    assert np.array_equal(data.var['gc_bias'].values, np.array([0.0, 1.0]))