PGAP2 2.2__tar.gz → 2.3__tar.gz

{pgap2-2.2 → pgap2-2.3/PGAP2.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PGAP2
-Version: 2.2
+Version: 2.3
 Summary: PGAP2: a comprehensive pan-genome analysis pipeline for prokaryotic genomes
 Author-email: bucongfan <bucf@big.ac.cn>
 Maintainer-email: bucongfan <bucf@big.ac.cn>

{pgap2-2.2/PGAP2.egg-info → pgap2-2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PGAP2
-Version: 2.2
+Version: 2.3
 Summary: PGAP2: a comprehensive pan-genome analysis pipeline for prokaryotic genomes
 Author-email: bucongfan <bucf@big.ac.cn>
 Maintainer-email: bucongfan <bucf@big.ac.cn>

{pgap2-2.2 → pgap2-2.3}/pgap2/postprocess/stat.py

@@ -55,6 +55,40 @@ def core_gene_model(x, C0, a, C_inf):
 def pan_gene_model(x, k, gamma):
     return k * x**gamma
 
+#make sure that only finite values are used in the fitting process.
+def _drop_invalid_fit_points(x_data, y_data):
+    x_data = np.asarray(x_data, dtype=float)
+    y_data = np.asarray(y_data, dtype=float)
+    valid_points = np.isfinite(x_data) & np.isfinite(y_data)
+    return x_data[valid_points], y_data[valid_points]
+
+
+def _get_core_fit_initial_params(core_gene_means):
+    if len(core_gene_means) == 0:
+        raise ValueError("No valid core gene points are available for curve fitting.")
+    C_inf = max(float(core_gene_means[-1]), 0.0)
+    C0 = max(float(core_gene_means[0]) - C_inf, 0.0)
+    return [C0, 0.05, C_inf]
+
+
+def _get_pan_fit_initial_params(pan_gene_means):
+    if len(pan_gene_means) == 0:
+        raise ValueError("No valid pan gene points are available for curve fitting.")
+    k = max(float(pan_gene_means[0]), 0.0)
+    return [k, 0.5]
+
+
+def _fit_pan_gene_loglog(x_data, pan_gene_means):
+    positive_points = (x_data > 0) & (pan_gene_means > 0)
+    if np.count_nonzero(positive_points) < 2:
+        raise RuntimeError(
+            "At least two positive points are required for log-log pan gene fitting.")
+
+    gamma, log_k = np.polyfit(
+        np.log(x_data[positive_points]), np.log(pan_gene_means[positive_points]), 1)
+    k = np.exp(log_k)
+    return np.asarray([k, gamma], dtype=float)
+
 
 def stat_pan_core(sampling_list, pav, i_strain, strain_num):
     core_list = []
@@ -122,7 +156,7 @@ def log_pan_genome_state(popt_pan, ofh):
 
 def fit_rerefaction(pan_rarefaction, outdir):
 
-    x_data = pan_rarefaction[0]
+    x_data = np.asarray(pan_rarefaction[0], dtype=float)
     core_gene_samples = pan_rarefaction[1]
     pan_gene_samples = pan_rarefaction[2]
 
@@ -134,27 +168,75 @@ def fit_rerefaction(pan_rarefaction, outdir):
             pan_gene_samples[i] = [
                 int(_) for _ in pan_gene_samples[i] if _ != 'NA']
 
-    core_gene_means = [np.mean(samples) for samples in core_gene_samples]
-    pan_gene_means = [np.mean(samples) for samples in pan_gene_samples]
-
-    ofh = open(file=f'{outdir}/postprocess.curve_fit.txt', mode='w')
-
-    logger.info("Fitting core gene data.")
-    popt_core, _ = curve_fit(core_gene_model, x_data,
-                             core_gene_means, p0=[2000, 0.1, 500])
-    logger.info(f"Optimal parameters for core genes: {popt_core}")
-    ofh.write(f'Optimal parameters for core genes: {popt_core}\n')
+    core_gene_means = np.asarray(
+        [np.mean(samples) for samples in core_gene_samples], dtype=float)
+    pan_gene_means = np.asarray(
+        [np.mean(samples) for samples in pan_gene_samples], dtype=float)
+
+    core_x_data, core_gene_means = _drop_invalid_fit_points(
+        x_data, core_gene_means)
+    pan_x_data, pan_gene_means = _drop_invalid_fit_points(
+        x_data, pan_gene_means)
+
+    with open(file=f'{outdir}/postprocess.curve_fit.txt', mode='w') as ofh:
+
+        logger.info("Fitting core gene data.")
+        try:
+            popt_core, _ = curve_fit(
+                core_gene_model,
+                core_x_data,
+                core_gene_means,
+                p0=_get_core_fit_initial_params(core_gene_means),
+                bounds=([0, 0, 0], [np.inf, np.inf, np.inf]),
+                maxfev=10000,
+            )
+            logger.info(f"Optimal parameters for core genes: {popt_core}")
+            ofh.write(f'Optimal parameters for core genes: {popt_core}\n')
+        except (RuntimeError, ValueError) as err:
+            popt_core = None
+            logger.warning(f"Core gene curve fitting failed: {err}")
+            ofh.write('Optimal parameters for core genes: NA\n')
+
+        logger.info("Fitting pan gene data.")
+        try:
+            popt_pan, _ = curve_fit(
+                pan_gene_model,
+                pan_x_data,
+                pan_gene_means,
+                p0=_get_pan_fit_initial_params(pan_gene_means),
+                bounds=([0, 0], [np.inf, np.inf]),
+                maxfev=10000,
+            )
+            logger.info(f"Optimal parameters for pan genes: {popt_pan}")
+            ofh.write(f'Optimal parameters for pan genes: {popt_pan}\n')
+        except (RuntimeError, ValueError) as err:
+            logger.warning(
+                f"Pan gene curve fitting failed with curve_fit: {err}. Trying log-log fallback.")
+            try:
+                popt_pan = _fit_pan_gene_loglog(pan_x_data, pan_gene_means)
+                logger.info(
+                    f"Optimal parameters for pan genes from log-log fallback: {popt_pan}")
+                ofh.write(
+                    f'Optimal parameters for pan genes: {popt_pan} (log-log fallback)\n')
+            except (RuntimeError, ValueError) as fallback_err:
+                popt_pan = None
+                logger.warning(
+                    f"Pan gene log-log fallback fitting failed: {fallback_err}")
+                ofh.write('Optimal parameters for pan genes: NA\n')
+
+        if popt_core is None:
+            ofh.write('Core Gene Fit: NA\n')
+        else:
+            ofh.write(
+                f'Core Gene Fit: y={popt_core[0]:.3f}*exp(-{popt_core[1]:.3f}*x)+ {popt_core[2]:.3f}\n')
 
-    logger.info("Fitting pan gene data.")
-    popt_pan, _ = curve_fit(pan_gene_model, x_data,
-                            pan_gene_means, p0=[1000, 0.8])
-    logger.info(f"Optimal parameters for pan genes: {popt_pan}")
-    ofh.write(f'Optimal parameters for pan genes: {popt_pan}\n')
-    ofh.write(
-        f'Core Gene Fit: y={popt_core[0]:.3f}*exp({popt_core[1]:.3f}*x)+ {popt_core[2]:.3f}\n')
-    ofh.write(
-        f'Pan Gene Fit: y={popt_pan[0]:.3f}*x^{popt_pan[1]:.3f} \n')
-    log_pan_genome_state(popt_pan, ofh)
+        if popt_pan is None:
+            ofh.write('Pan Gene Fit: NA\n')
+            ofh.write('Pan-genome state: NA\n')
+        else:
+            ofh.write(
+                f'Pan Gene Fit: y={popt_pan[0]:.3f}*x^{popt_pan[1]:.3f} \n')
+            log_pan_genome_state(popt_pan, ofh)
 
     # plt.figure(figsize=(8, 6))
 

{pgap2-2.2 → pgap2-2.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "PGAP2"
-version = "2.2"
+version = "2.3"
 dynamic = ["readme"]
 keywords = ["pan-genome", "prokaryotic", "genomes", "pipeline","cluster"]
 

pgap2-2.3/tests/test_post.py

@@ -0,0 +1,95 @@
+import pytest
+import numpy as np
+
+
+@pytest.fixture
+def test_basic():
+    basic_file = 'tests/output/basic.pkl'
+    from pgap2.lib.pklcheck import PklCheck
+    from pgap2.lib.basic import Basic
+    import pickle
+
+    with open(f'{basic_file}', 'rb') as fh:
+        previous: PklCheck = pickle.load(fh)
+        decode_status = previous.decode()
+        assert decode_status is True, "decode did not return True for matching parameters"
+        basic: Basic = previous.data_dump('basic')
+        assert isinstance(
+            basic, Basic), "Basic object was not created correctly"
+        return basic
+
+
+def test_get_pan_group(test_basic):
+    from pgap2.postprocess.stat import get_pan_group
+    basic = test_basic
+    outdir = 'tests/output'
+    basic.load_pav(file=f'{outdir}/pgap2.partition.gene_content.pav')
+    pan_group_freq, pan_para_stat = get_pan_group(basic.pav, outdir)
+    assert len(
+        pan_group_freq) == 3, "get_pan_group did not return the expected number of groups"
+    assert len(
+        pan_para_stat) == 5, "get_pan_group did not return the expected number of parameters"
+
+
+def test_rare(test_basic):
+    from pgap2.postprocess.stat import get_rarefaction, fit_rerefaction
+    basic = test_basic
+    outdir = 'tests/output/'
+    basic.load_pav(file=f'{outdir}/pgap2.partition.gene_content.pav')
+    pan_profile, new_clusters = get_rarefaction(
+        basic.pav, outdir, N=1)
+    assert len(
+        pan_profile) == 3, "get_rarefaction did not return the expected number of profiles"
+    assert len(
+        new_clusters) == 2, "get_rarefaction did not return the expected number of clusters"
+
+
+def test_fit_rarefaction_keeps_running_when_core_fit_fails(tmp_path, monkeypatch):
+    from pgap2.postprocess import stat
+
+    pan_profile = [
+        [1, 2, 3, 4],
+        [[3000], [2500], [2300], [2200]],
+        [[4000], [4500], [4800], [5100]],
+    ]
+    original_curve_fit = stat.curve_fit
+
+    def fail_core_fit(model, *args, **kwargs):
+        if model is stat.core_gene_model:
+            raise RuntimeError("forced core failure")
+        return original_curve_fit(model, *args, **kwargs)
+
+    monkeypatch.setattr(stat, "curve_fit", fail_core_fit)
+
+    stat.fit_rerefaction(pan_profile, tmp_path)
+
+    content = (tmp_path / "postprocess.curve_fit.txt").read_text()
+    assert "Optimal parameters for core genes: NA" in content
+    assert "Pan Gene Fit: y=" in content
+
+
+def test_fit_rarefaction_uses_pan_loglog_fallback(tmp_path, monkeypatch):
+    from pgap2.postprocess import stat
+
+    pan_profile = [
+        [1, 2, 4, 8],
+        [[3000], [2500], [2300], [2200]],
+        [[1000], [1414], [2000], [2828]],
+    ]
+
+    def fail_pan_fit(model, *args, **kwargs):
+        if model is stat.pan_gene_model:
+            raise RuntimeError("forced pan failure")
+        return (
+            np.asarray([800.0, 0.1, 2200.0], dtype=float),
+            np.eye(3),
+        )
+
+    monkeypatch.setattr(stat, "curve_fit", fail_pan_fit)
+
+    stat.fit_rerefaction(pan_profile, tmp_path)
+
+    content = (tmp_path / "postprocess.curve_fit.txt").read_text()
+    assert "Optimal parameters for pan genes:" in content
+    assert "log-log fallback" in content
+    assert "Pan Gene Fit: y=" in content

pgap2-2.2/tests/test_post.py

@@ -1,43 +0,0 @@
-import pytest
-
-
-@pytest.fixture
-def test_basic():
-    basic_file = 'tests/output/basic.pkl'
-    from pgap2.lib.pklcheck import PklCheck
-    from pgap2.lib.basic import Basic
-    import pickle
-
-    with open(f'{basic_file}', 'rb') as fh:
-        previous: PklCheck = pickle.load(fh)
-        decode_status = previous.decode()
-        assert decode_status is True, "decode did not return True for matching parameters"
-        basic: Basic = previous.data_dump('basic')
-        assert isinstance(
-            basic, Basic), "Basic object was not created correctly"
-        return basic
-
-
-def test_get_pan_group(test_basic):
-    from pgap2.postprocess.stat import get_pan_group
-    basic = test_basic
-    outdir = 'tests/output'
-    basic.load_pav(file=f'{outdir}/pgap2.partition.gene_content.pav')
-    pan_group_freq, pan_para_stat = get_pan_group(basic.pav, outdir)
-    assert len(
-        pan_group_freq) == 3, "get_pan_group did not return the expected number of groups"
-    assert len(
-        pan_para_stat) == 5, "get_pan_group did not return the expected number of parameters"
-
-
-def test_rare(test_basic):
-    from pgap2.postprocess.stat import get_rarefaction, fit_rerefaction
-    basic = test_basic
-    outdir = 'tests/output/'
-    basic.load_pav(file=f'{outdir}/pgap2.partition.gene_content.pav')
-    pan_profile, new_clusters = get_rarefaction(
-        basic.pav, outdir, N=1)
-    assert len(
-        pan_profile) == 3, "get_rarefaction did not return the expected number of profiles"
-    assert len(
-        new_clusters) == 2, "get_rarefaction did not return the expected number of clusters"