csa3d 0.2.0__tar.gz → 0.2.2__tar.gz

{csa3d-0.2.0 → csa3d-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: csa3d
-Version: 0.2.0
+Version: 0.2.2
 Summary: Extract and analyze enzyme catalytic sites in 3D from the Mechanism and Catalytic Site Atlas (M-CSA)
 Author-email: Ioannis Riziotis <riziotis@ebi.ac.uk>
 License: Apache-2.0
@@ -30,9 +30,10 @@ Requires-Dist: scipy>=1.7.0
 Requires-Dist: scikit-learn>=0.24.0
 Requires-Dist: rdkit>=2021.09.0
 Requires-Dist: natsort>=8.0.0
-Requires-Dist: dendrogram-cut>=0.1.0
-Requires-Dist: parity-core>=0.5.0
 Requires-Dist: rmsd>=1.4.0
+Provides-Extra: clustering
+Requires-Dist: matplotlib>=3.5; extra == "clustering"
+Requires-Dist: plotly>=5.0; extra == "clustering"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: pytest-cov>=4.0; extra == "dev"
@@ -105,28 +106,43 @@ pip install -e ".[dev,docs]"
 
 ## Quick Start
 
-### 1. Configure Data Directory
+### 1. Automatic Data Setup (Recommended)
 
-First, you need to set up the M-CSA data files. You can either:
+Download and setup M-CSA data automatically:
 
-**Option A: Download automatically (recommended)**
 ```python
-from csa3d import config
-config.setup_data("/path/to/csa3d_data")  # Downloads ~50 GB
+from csa3d import setup_data
+
+# Downloads ~50 GB of data to ~/.csa3d/data
+# Set up specific location with:
+# setup_data("/path/to/csa3d_data")
+success = setup_data()
+
+if success:
+    print("✓ Data setup complete!")
 ```
 
-**Option B: Set environment variable**
+Or via command line:
+```bash
+python -m csa3d.data_setup /path/to/csa3d_data
+```
+
+### 2. Manual Data Configuration
+
+If you prefer to download data manually:
+
+**Option A: Environment variable**
 ```bash
 export CSA3D_DATA_DIR="/path/to/csa3d_data"
 ```
 
-**Option C: Set in code**
+**Option B: Set in code**
 ```python
-from csa3d import config
-config.set_data_dir("/path/to/csa3d_data")
+from csa3d import set_data_dir
+set_data_dir("/path/to/csa3d_data")
 ```
 
-### 2. Load and Inspect Active Sites
+### 3. Load and Inspect Active Sites
 
 ```python
 from csa3d import Mcsa
@@ -151,7 +167,7 @@ for entry in db:
         print(f"    Resolution: {site.resolution} Å")
 ```
 
-### 3. Clustering and Template Creation
+### 4. Clustering and Template Creation
 
 ```python
 # Build multiple entries

{csa3d-0.2.0 → csa3d-0.2.2}/README.md

@@ -51,28 +51,43 @@ pip install -e ".[dev,docs]"
 
 ## Quick Start
 
-### 1. Configure Data Directory
+### 1. Automatic Data Setup (Recommended)
 
-First, you need to set up the M-CSA data files. You can either:
+Download and setup M-CSA data automatically:
 
-**Option A: Download automatically (recommended)**
 ```python
-from csa3d import config
-config.setup_data("/path/to/csa3d_data")  # Downloads ~50 GB
+from csa3d import setup_data
+
+# Downloads ~50 GB of data to ~/.csa3d/data
+# Set up specific location with:
+# setup_data("/path/to/csa3d_data")
+success = setup_data()
+
+if success:
+    print("✓ Data setup complete!")
 ```
 
-**Option B: Set environment variable**
+Or via command line:
+```bash
+python -m csa3d.data_setup /path/to/csa3d_data
+```
+
+### 2. Manual Data Configuration
+
+If you prefer to download data manually:
+
+**Option A: Environment variable**
 ```bash
 export CSA3D_DATA_DIR="/path/to/csa3d_data"
 ```
 
-**Option C: Set in code**
+**Option B: Set in code**
 ```python
-from csa3d import config
-config.set_data_dir("/path/to/csa3d_data")
+from csa3d import set_data_dir
+set_data_dir("/path/to/csa3d_data")
 ```
 
-### 2. Load and Inspect Active Sites
+### 3. Load and Inspect Active Sites
 
 ```python
 from csa3d import Mcsa
@@ -97,7 +112,7 @@ for entry in db:
         print(f"    Resolution: {site.resolution} Å")
 ```
 
-### 3. Clustering and Template Creation
+### 4. Clustering and Template Creation
 
 ```python
 # Build multiple entries

csa3d-0.2.2/csa3d/__init__.py

@@ -0,0 +1,12 @@
+"""CSA-3D: Extract and analyze enzyme catalytic sites from M-CSA."""
+
+from .data_setup import setup_data
+from .config import set_data_dir, get_data_dir
+
+__version__ = "0.2.2"
+
+__all__ = [
+    "setup_data",
+    "set_data_dir",
+    "get_data_dir",
+]

csa3d-0.2.2/csa3d/_vendor/__init__.py

@@ -0,0 +1,7 @@
+"""Vendored dependencies for CSA-3D.
+
+This package contains minimal implementations of research-specific dependencies
+that are not available on PyPI.
+"""
+
+__all__ = ['dendrogram_cut']

csa3d-0.2.2/csa3d/_vendor/dendrogram_cut.py

@@ -0,0 +1,469 @@
+import itertools
+
+import numpy as np
+import scipy.spatial
+import scipy.cluster
+import matplotlib.cm
+import plotly.graph_objs as go
+from sklearn.metrics import silhouette_score
+
+
+class DendrogramCut:
+    def __init__(self, k_max, method='average'):
+        self.k_max = k_max
+        self.method = method
+
+    def fit(self, distance_matrix):
+        '''
+        Build linkage_stats
+            css: cross sum of square when merging c1 and c2
+            tss: total sum of square of merged cluster
+        '''
+        # Handle both condensed and square distance matrices
+        if distance_matrix.ndim == 1:
+            self.distance_matrix = scipy.spatial.distance.squareform(distance_matrix)
+            self.linkage = scipy.cluster.hierarchy.linkage(distance_matrix, method=self.method, optimal_ordering=True)
+        else:
+            self.distance_matrix = distance_matrix
+            self.linkage = scipy.cluster.hierarchy.linkage(scipy.spatial.distance.pdist(distance_matrix), method=self.method, optimal_ordering=True)
+
+        self.n_data = self.distance_matrix.shape[0]
+        self.linkage_stats = [{'c1': 0, 'c2': 0, 'css': 0, 'tss': 0, 'indices': set()} for _ in range(2 * self.n_data - 1)]
+
+        for i in range(self.n_data):
+            self.linkage_stats[i]['c1'] = i
+            self.linkage_stats[i]['c2'] = i
+            self.linkage_stats[i]['indices'].add(i)
+
+        for i, (c1, c2, _, _) in enumerate(self.linkage):
+            c1 = int(c1)
+            c2 = int(c2)
+            self.linkage_stats[i + self.n_data]['c1'] = c1
+            self.linkage_stats[i + self.n_data]['c2'] = c2
+            self.linkage_stats[i + self.n_data]['indices'].update(self.linkage_stats[c1]['indices'])
+            self.linkage_stats[i + self.n_data]['indices'].update(self.linkage_stats[c2]['indices'])
+
+        for i in range(self.n_data, 2 * self.n_data - 1):
+            c1 = self.linkage_stats[i]['c1']
+            c2 = self.linkage_stats[i]['c2']
+            c1_indices = np.asarray(list(self.linkage_stats[c1]['indices']))
+            c2_indices = np.asarray(list(self.linkage_stats[c2]['indices']))
+
+            sample_distances = self.distance_matrix[c1_indices, :][:, c2_indices]
+            self.linkage_stats[i]['css'] = np.sum(sample_distances ** 2)
+            self.linkage_stats[i]['tss'] = self.linkage_stats[i]['css'] + self.linkage_stats[c1]['tss'] + self.linkage_stats[c2]['tss']
+
+        ### dynamic programming ###
+        '''
+        Dynamic programming
+            kl_mat[i, k]: the number of clusters in the left branch of linkage_stat[$i], maximal cluster $k
+            mss_mat[i, k]: the optimal mean square error achieved at linkage_stat[$i], maximal cluster $k
+        '''
+        self.kl_mat = np.zeros((self.n_data * 2 - 1, self.k_max + 1), dtype=int)
+        self.mss_mat = np.zeros((self.n_data * 2 - 1, self.k_max + 1), dtype=float) + np.inf
+
+        for i in range(self.n_data):
+            self.mss_mat[i, 1] = 0
+
+        for i in range(self.n_data, 2 * self.n_data - 1):
+            self.mss_mat[i, 1] = self.linkage_stats[i]['tss'] / len(self.linkage_stats[i]['indices'])
+
+        for i in range(self.n_data, 2 * self.n_data - 1):
+            for k in range(2, self.k_max + 1):
+                kl_min = 0
+                mss_min = np.inf
+                for kl in range(1, k):
+                    tss = self.mss_mat[self.linkage_stats[i]['c1'], kl] + self.mss_mat[self.linkage_stats[i]['c2'], k - kl]
+                    if tss < mss_min:
+                        kl_min = kl
+                        mss_min = tss
+                self.kl_mat[i, k] = kl_min
+                self.mss_mat[i, k] = mss_min
+
+        return self
+
+    def _get_cut_nodes(self, v, k):
+        if k == 1:
+            yield v
+        else:
+            yield from self._get_cut_nodes(self.linkage_stats[v]['c1'], self.kl_mat[v, k])
+            yield from self._get_cut_nodes(self.linkage_stats[v]['c2'], k - self.kl_mat[v, k])
+
+    def get_cluster_mss(self, k):
+        total_mss = 0.
+        for cid in self._get_cut_nodes(2 * self.n_data - 2, k):
+            total_mss += self.linkage_stats[cid]['tss'] / len(self.linkage_stats[cid]['indices'])
+
+        return total_mss
+
+    def get_cluster_label(self, k):
+        ### get flat clusters ###
+        z = np.zeros(self.n_data, dtype=int) - 1
+        for c, cid in enumerate(self._get_cut_nodes(2 * self.n_data - 2, k)):
+            for i in self.linkage_stats[cid]['indices']:
+                if z[i] != -1:
+                    print(i)
+                z[i] = c
+
+        return z
+
+    def _dirichlet_process_kl(self, n_list, alpha_):
+        out = 0.
+        count = alpha_
+        for n in n_list:
+            out -= np.log(alpha_ / count)
+            count += 1
+            for i in range(1, n):
+                out -= np.log(i / count)
+                count += 1
+        return out
+
+    def pac_bayesian_cut(self, alpha_=1., lambda_=1.):
+        min_loss = np.inf
+        min_loss_k = None
+
+        for k in range(1, self.k_max + 1):
+            n_list = [len(self.linkage_stats[c]['indices']) for c in self._get_cut_nodes(2 * self.n_data - 2, k)]
+            total_mss = self.get_cluster_mss(k)
+            loss = total_mss + self._dirichlet_process_kl(n_list, alpha_) / lambda_
+            if loss < min_loss:
+                min_loss = loss
+                min_loss_k = k
+
+        return min_loss_k
+
+    def dendrogram_plot(self, k=None, label=None):
+        cmap = matplotlib.cm.get_cmap('viridis')
+
+        order = scipy.cluster.hierarchy.leaves_list(self.linkage)
+
+        dendrogram_stat = dict((v, {'x': i, 'y': 0}) for i, v in enumerate(order))
+
+        data = []
+
+        ### plot dendrogram ###
+        if k is None:
+            norm = np.max(self.linkage[:, 2])
+
+            for res in self.linkage:
+                new_cluster = len(dendrogram_stat)
+                dendrogram_stat[new_cluster] = {'x': (dendrogram_stat[res[0]]['x'] + dendrogram_stat[res[1]]['x']) / 2, 'y': res[2]}
+                x_range = np.array([dendrogram_stat[res[0]]['x'], dendrogram_stat[res[0]]['x'], dendrogram_stat[res[1]]['x'], dendrogram_stat[res[1]]['x']])
+                y_range = np.array([dendrogram_stat[res[0]]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[res[1]]['y']])
+
+                data.append(
+                    {
+                        'hoverinfo': 'text',
+                        'marker': {'color': 'rgb({}, {}, {})'.format(*cmap(res[2] / norm, bytes=True)[:3])},
+                        'mode': 'lines',
+                        'type': 'scatter',
+                        'x': x_range,
+                        'xaxis': 'x',
+                        'y': y_range,
+                        'yaxis': 'y'
+                    }
+                )
+        else:
+            pivot = set(self._get_cut_nodes(2 * self.n_data - 2, k))
+            z = self.get_cluster_label(k)
+            colorscale = dict((i, z[i] / (k - 1)) for i in range(self.n_data))
+
+            for res in self.linkage:
+                new_cluster = len(dendrogram_stat)
+                colorscale[new_cluster] = (colorscale[int(res[0])] + colorscale[int(res[1])]) / 2
+                dendrogram_stat[new_cluster] = {'x': (dendrogram_stat[res[0]]['x'] + dendrogram_stat[res[1]]['x']) / 2, 'y': res[2]}
+                x_range = np.array([dendrogram_stat[res[0]]['x'], dendrogram_stat[res[0]]['x'], dendrogram_stat[res[1]]['x'], dendrogram_stat[res[1]]['x']])
+                y_range = np.array([dendrogram_stat[res[0]]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[res[1]]['y']])
+
+                if new_cluster in pivot:
+                    data.append(
+                        {
+                            'x': [dendrogram_stat[new_cluster]['x']],
+                            'y': [dendrogram_stat[new_cluster]['y']],
+                            'type': 'scatter',
+                            'mode': 'markers',
+                            'xaxis': 'x',
+                            'yaxis': 'y',
+                            'marker': {'size': [20], 'color': 'rgb({}, {}, {})'.format(*cmap(colorscale[new_cluster], bytes=True)[:3])}
+                        }
+                    )
+
+                data.append(
+                    {
+                        'hoverinfo': 'text',
+                        'marker': {'color': 'rgb({}, {}, {})'.format(*cmap(colorscale[new_cluster], bytes=True)[:3])},
+                        'mode': 'lines',
+                        'type': 'scatter',
+                        'x': x_range,
+                        'xaxis': 'x',
+                        'y': y_range,
+                        'yaxis': 'y'
+                    }
+                )
+
+        fig = go.Figure(data=data)
+
+        ### edit layout ###
+        if label is not None:
+            fig.update_layout(
+                {
+                    'xaxis': {
+                        'tickmode': 'array',
+                        'tickvals': np.arange(len(label)),
+                        'ticktext': label[order]
+                    }
+                }
+            )
+
+        fig.update_layout(
+            {
+                'width': 800,
+                'height': 800,
+                'showlegend': False,
+                'hovermode': 'closest',
+                'xaxis': {
+                    'domain': [0, 1],
+                    'rangemode': 'tozero',
+                    'mirror': False,
+                    'showgrid': False,
+                    'showline': False,
+                    'zeroline': False,
+                    'showticklabels': True,
+                    'ticks': ""
+                },
+                'yaxis': {
+                    'domain': [0, 1],
+                    'rangemode': 'tozero',
+                    'mirror': False,
+                    'showgrid': False,
+                    'showline': False,
+                    'zeroline': False,
+                    'showticklabels': True,
+                    'ticks': ""
+                },
+                'paper_bgcolor': 'rgba(0,0,0,0)',
+                'plot_bgcolor': 'rgba(0,0,0,0)'
+            }
+        )
+
+        fig.show()
+
+    def heatmap_with_dendrogram_plot(self, k=None, label=None):
+        cmap = matplotlib.cm.get_cmap('viridis')
+
+        order = scipy.cluster.hierarchy.leaves_list(self.linkage)
+
+        dendrogram_stat = dict((v, {'x': i, 'y': 0}) for i, v in enumerate(order))
+
+        data = []
+
+        ### plot dendrogram ###
+        if k is None:
+            norm = np.max(self.linkage[:, 2])
+
+            for res in self.linkage:
+                new_cluster = len(dendrogram_stat)
+                dendrogram_stat[new_cluster] = {'x': (dendrogram_stat[res[0]]['x'] + dendrogram_stat[res[1]]['x']) / 2, 'y': res[2]}
+                x_range = np.array([dendrogram_stat[res[0]]['x'], dendrogram_stat[res[0]]['x'], dendrogram_stat[res[1]]['x'], dendrogram_stat[res[1]]['x']])
+                y_range = np.array([dendrogram_stat[res[0]]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[res[1]]['y']])
+
+                data.append(
+                    {
+                        'hoverinfo': 'text',
+                        'marker': {'color': 'rgb({}, {}, {})'.format(*cmap(res[2] / norm, bytes=True)[:3])},
+                        'mode': 'lines',
+                        'type': 'scatter',
+                        'x': x_range,
+                        'xaxis': 'x',
+                        'y': y_range,
+                        'yaxis': 'y2'
+                    }
+                )
+
+                data.append(
+                    {
+                        'hoverinfo': 'text',
+                        'marker': {'color': 'rgb({}, {}, {})'.format(*cmap(res[2] / norm, bytes=True)[:3])},
+                        'mode': 'lines',
+                        'type': 'scatter',
+                        'x': -y_range,
+                        'xaxis': 'x2',
+                        'y': x_range,
+                        'yaxis': 'y'
+                    }
+                )
+        else:
+            pivot = set(self._get_cut_nodes(2 * self.n_data - 2, k))
+            z = self.get_cluster_label(k)
+            colorscale = dict((i, z[i] / (k - 1)) for i in range(self.n_data))
+
+            for res in self.linkage:
+                new_cluster = len(dendrogram_stat)
+                colorscale[new_cluster] = (colorscale[int(res[0])] + colorscale[int(res[1])]) / 2
+                dendrogram_stat[new_cluster] = {'x': (dendrogram_stat[res[0]]['x'] + dendrogram_stat[res[1]]['x']) / 2, 'y': res[2]}
+                x_range = np.array([dendrogram_stat[res[0]]['x'], dendrogram_stat[res[0]]['x'], dendrogram_stat[res[1]]['x'], dendrogram_stat[res[1]]['x']])
+                y_range = np.array([dendrogram_stat[res[0]]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[new_cluster]['y'], dendrogram_stat[res[1]]['y']])
+
+                if new_cluster in pivot:
+                    data.append(
+                        {
+                            'x': [dendrogram_stat[new_cluster]['x']],
+                            'y': [dendrogram_stat[new_cluster]['y']],
+                            'type': 'scatter',
+                            'mode': 'markers',
+                            'xaxis': 'x',
+                            'yaxis': 'y2',
+                            'marker': {'size': [10], 'color': 'rgb({}, {}, {})'.format(*cmap(colorscale[new_cluster], bytes=True)[:3])}
+                        }
+                    )
+                    data.append(
+                        {
+                            'x': [-dendrogram_stat[new_cluster]['y']],
+                            'y': [dendrogram_stat[new_cluster]['x']],
+                            'type': 'scatter',
+                            'mode': 'markers',
+                            'xaxis': 'x2',
+                            'yaxis': 'y',
+                            'marker': {'size': [10], 'color': 'rgb({}, {}, {})'.format(*cmap(colorscale[new_cluster], bytes=True)[:3])}
+                        }
+                    )
+
+                data.append(
+                    {
+                        'hoverinfo': 'text',
+                        'marker': {'color': 'rgb({}, {}, {})'.format(*cmap(colorscale[new_cluster], bytes=True)[:3])},
+                        'mode': 'lines',
+                        'type': 'scatter',
+                        'x': x_range,
+                        'xaxis': 'x',
+                        'y': y_range,
+                        'yaxis': 'y2'
+                    }
+                )
+
+                data.append(
+                    {
+                        'hoverinfo': 'text',
+                        'marker': {'color': 'rgb({}, {}, {})'.format(*cmap(colorscale[new_cluster], bytes=True)[:3])},
+                        'mode': 'lines',
+                        'type': 'scatter',
+                        'x': -y_range,
+                        'xaxis': 'x2',
+                        'y': x_range,
+                        'yaxis': 'y'
+                    }
+                )
+
+            ### plot hard cluster region ###
+
+            ordered_z = z[order]
+            block_size = [sum(1 for __ in g) for _, g in itertools.groupby(ordered_z)]
+            start = -0.5
+            for i in range(len(block_size)):
+                data.append(
+                    go.Scatter(
+                        x=[start, start, start+block_size[i], start+block_size[i], start],
+                        y=[start, start+block_size[i], start+block_size[i], start, start],
+                        mode='lines',
+                        hoverinfo='none',
+                        line=dict(color='#FF0000')
+                    )
+                )
+                start += block_size[i]
+
+        ### plot heatmap ###
+        data.append(
+            go.Heatmap(
+                z=self.distance_matrix[order, :][:, order],
+                colorscale='Blues',
+                reversescale=True,
+                hoverinfo='x+y+z'
+            )
+        )
+
+        fig = go.Figure(data=data)
+
+        ### edit layout ###
+        if label is not None:
+            fig.update_layout(
+                {
+                    'xaxis': {
+                        'tickmode': 'array',
+                        'tickvals': np.arange(len(label)),
+                        'ticktext': label[order]
+                    },
+                    'yaxis': {
+                        'tickmode': 'array',
+                        'tickvals': np.arange(len(label)),
+                        'ticktext': label[order],
+                        'autorange': 'reversed'
+                    }
+                }
+            )
+
+        fig.update_layout(
+            {
+                'width': 800,
+                'height': 800,
+                'showlegend': False,
+                'hovermode': 'closest',
+                'xaxis': {
+                    'domain': [.15, 1],
+                    'rangemode': 'tozero',
+                    'mirror': False,
+                    'showgrid': False,
+                    'showline': False,
+                    'zeroline': False,
+                    'showticklabels': False,
+                    'ticks': ""
+                },
+                'xaxis2': {
+                    'domain': [0, .15],
+                    'rangemode': 'tozero',
+                    'mirror': False,
+                    'showgrid': False,
+                    'showline': False,
+                    'zeroline': False,
+                    'showticklabels': False,
+                    'ticks': ""
+                },
+                'yaxis': {
+                    'domain': [0, .85],
+                    'rangemode': 'tozero',
+                    'mirror': False,
+                    'showgrid': False,
+                    'showline': False,
+                    'zeroline': False,
+                    'showticklabels': False,
+                    'ticks': ""
+                },
+                'yaxis2': {
+                    'domain':[.825, .975],
+                    'rangemode': 'tozero',
+                    'mirror': False,
+                    'showgrid': False,
+                    'showline': False,
+                    'zeroline': False,
+                    'showticklabels': False,
+                    'ticks': ""
+                }
+            }
+        )
+
+        fig.show()
+
+    def elbow_plot(self):
+        x = np.arange(self.k_max) + 1
+
+        y = np.asarray([sum(self.linkage_stats[i]['tss'] / len(self.linkage_stats[i]['indices']) for i in self._get_cut_nodes(-1, k)) for k in range(1, self.k_max + 1)])
+        y = (y[0] - y) / y[0]
+
+        fig = go.Figure(data=[go.Scatter(x=x, y=y, mode='lines')])
+        fig.show()
+
+    def silhouette_score_plot(self):
+        x = np.arange(2, self.k_max + 1)
+        y = [silhouette_score(self.distance_matrix, self.get_cluster_label(k), metric='precomputed') for k in x]
+
+        fig = go.Figure(data=go.Scatter(x=x, y=y, mode='lines'))
+        fig.show()

csa3d-0.2.2/csa3d/data_setup.py

@@ -0,0 +1,224 @@
+"""Automatic data setup and download for CSA-3D.
+
+Downloads and organizes required data files from official sources.
+"""
+
+import logging
+import os
+from pathlib import Path
+from typing import Optional
+import urllib.request
+import urllib.error
+from urllib.parse import urljoin
+
+logger = logging.getLogger(__name__)
+
+
+class DataDownloader:
+    """Download and setup CSA-3D data files."""
+
+    # Data source URLs
+    SIFTS_URL = "https://www.ebi.ac.uk/pdbe/download/sifts/csv/"
+    KEGG_URL = "https://rest.kegg.jp/get/"
+    RHEA_URL = "https://www.rhea-db.org/download/rhea/"
+    PDBE_URL = "https://www.ebi.ac.uk/pdbe/download/"
+    MCSA_URL = "https://www.ebi.ac.uk/thornton-srv/m-csa/download/"
+
+    def __init__(self, data_dir: Path):
+        """Initialize data downloader.
+
+        Args:
+            data_dir: Root directory for CSA-3D data
+        """
+        self.data_dir = Path(data_dir)
+        self.data_dir.mkdir(parents=True, exist_ok=True)
+
+    def download_file(self, url: str, output_path: Path) -> bool:
+        """Download a file from URL.
+
+        Args:
+            url: URL to download from
+            output_path: Path to save file
+
+        Returns:
+            True if successful, False otherwise
+        """
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+
+        try:
+            logger.info(f"Downloading {url} → {output_path}")
+            urllib.request.urlretrieve(url, output_path)
+            logger.info(f"✓ Downloaded {output_path.name}")
+            return True
+        except urllib.error.URLError as e:
+            logger.warning(f"✗ Failed to download {url}: {e}")
+            return False
+        except Exception as e:
+            logger.error(f"Unexpected error downloading {url}: {e}")
+            return False
+
+    def setup_sifts(self) -> bool:
+        """Setup SIFTS (PDB-UniProt mappings).
+
+        Downloads from EBI PDBe SIFTS data.
+
+        Returns:
+            True if successful
+        """
+        logger.info("Setting up SIFTS data...")
+        sifts_dir = self.data_dir / "sifts"
+        sifts_dir.mkdir(parents=True, exist_ok=True)
+
+        files = {
+            "uniprot_pdb.csv": "uniprot_pdb.csv",
+            "pdb_chain_uniprot.csv": "pdb_chain_uniprot.csv",
+            "pdb_chain_enzyme.csv": "pdb_chain_enzyme.csv",
+        }
+
+        success = True
+        for filename, remote_name in files.items():
+            url = urljoin(self.SIFTS_URL, remote_name)
+            output = sifts_dir / filename
+            if not self.download_file(url, output):
+                success = False
+
+        return success
+
+    def setup_kegg(self) -> bool:
+        """Setup KEGG reaction data.
+
+        Returns:
+            True if successful
+        """
+        logger.info("Setting up KEGG data...")
+        kegg_dir = self.data_dir / "kegg"
+        kegg_dir.mkdir(parents=True, exist_ok=True)
+
+        # KEGG reactions endpoint
+        output = kegg_dir / "kegg_reaction_components.csv"
+        logger.info("Note: KEGG data requires special API handling")
+        logger.info("Please obtain KEGG reaction data from: https://www.kegg.jp/")
+        return False
+
+    def setup_rhea(self) -> bool:
+        """Setup RHEA reaction data.
+
+        Downloads from RHEA database.
+
+        Returns:
+            True if successful
+        """
+        logger.info("Setting up RHEA data...")
+        rhea_dir = self.data_dir / "rhea"
+        rhea_dir.mkdir(parents=True, exist_ok=True)
+
+        url = urljoin(self.RHEA_URL, "rhea_reaction_components.csv")
+        output = rhea_dir / "rhea_reaction_components.csv"
+        return self.download_file(url, output)
+
+    def setup_pdbe(self) -> bool:
+        """Setup PDBe data (ions, cofactors, resolution).
+
+        Returns:
+            True if successful
+        """
+        logger.info("Setting up PDBe data...")
+        pdbe_dir = self.data_dir / "pdbe"
+        pdbe_dir.mkdir(parents=True, exist_ok=True)
+
+        files = {
+            "pdb_ions.csv": "pdb_ions.csv",
+            "pdb_cofactors.csv": "pdb_cofactors.csv",
+            "resolu.idx": "resolu.idx",
+        }
+
+        success = True
+        for filename, remote_name in files.items():
+            url = urljoin(self.PDBE_URL, remote_name)
+            output = pdbe_dir / filename
+            if not self.download_file(url, output):
+                success = False
+
+        return success
+
+    def setup_mcsa(self) -> bool:
+        """Setup M-CSA catalytic site data.
+
+        Downloads from M-CSA database.
+
+        Returns:
+            True if successful
+        """
+        logger.info("Setting up M-CSA data...")
+        mcsa_dir = self.data_dir / "mcsa"
+        mcsa_dir.mkdir(parents=True, exist_ok=True)
+
+        url = urljoin(self.MCSA_URL, "mcsa_cofactors.csv")
+        output = mcsa_dir / "mcsa_cofactors.csv"
+        return self.download_file(url, output)
+
+    def setup_all(self, skip_optional: bool = False) -> dict:
+        """Setup all data files.
+
+        Args:
+            skip_optional: Skip optional data sources (KEGG)
+
+        Returns:
+            Dictionary with setup status for each source
+        """
+        logger.info(f"Starting automatic CSA-3D data setup to {self.data_dir}")
+
+        results = {}
+        results["sifts"] = self.setup_sifts()
+        results["rhea"] = self.setup_rhea()
+        results["pdbe"] = self.setup_pdbe()
+        results["mcsa"] = self.setup_mcsa()
+
+        if not skip_optional:
+            results["kegg"] = self.setup_kegg()
+
+        # Summary
+        logger.info("\n" + "=" * 60)
+        logger.info("Data Setup Summary:")
+        for source, success in results.items():
+            status = "✓ Complete" if success else "✗ Failed/Skipped"
+            logger.info(f"  {source.upper()}: {status}")
+        logger.info("=" * 60)
+
+        return results
+
+
+def setup_data(data_dir: Optional[str] = None) -> bool:
+    """Setup CSA-3D data automatically.
+
+    Args:
+        data_dir: Directory for data files. Defaults to ~/.csa3d/data
+
+    Returns:
+        True if setup successful
+    """
+    if data_dir is None:
+        data_dir = str(Path.home() / ".csa3d" / "data")
+
+    logger.info(f"Setting up CSA-3D data in: {data_dir}")
+
+    downloader = DataDownloader(Path(data_dir))
+    results = downloader.setup_all()
+
+    success = all(results.values())
+    if success:
+        logger.info(f"\n✓ Data setup complete!")
+        logger.info(f"Set environment variable: export CSA3D_DATA_DIR={data_dir}")
+    else:
+        logger.warning("\n⚠ Some data sources failed to download")
+        logger.info("Please check logs above and retry")
+
+    return success
+
+
+if __name__ == "__main__":
+    logging.basicConfig(
+        level=logging.INFO,
+        format="%(levelname)s: %(message)s"
+    )
+    setup_data()

{csa3d-0.2.0 → csa3d-0.2.2}/csa3d.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: csa3d
-Version: 0.2.0
+Version: 0.2.2
 Summary: Extract and analyze enzyme catalytic sites in 3D from the Mechanism and Catalytic Site Atlas (M-CSA)
 Author-email: Ioannis Riziotis <riziotis@ebi.ac.uk>
 License: Apache-2.0
@@ -30,9 +30,10 @@ Requires-Dist: scipy>=1.7.0
 Requires-Dist: scikit-learn>=0.24.0
 Requires-Dist: rdkit>=2021.09.0
 Requires-Dist: natsort>=8.0.0
-Requires-Dist: dendrogram-cut>=0.1.0
-Requires-Dist: parity-core>=0.5.0
 Requires-Dist: rmsd>=1.4.0
+Provides-Extra: clustering
+Requires-Dist: matplotlib>=3.5; extra == "clustering"
+Requires-Dist: plotly>=5.0; extra == "clustering"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: pytest-cov>=4.0; extra == "dev"
@@ -105,28 +106,43 @@ pip install -e ".[dev,docs]"
 
 ## Quick Start
 
-### 1. Configure Data Directory
+### 1. Automatic Data Setup (Recommended)
 
-First, you need to set up the M-CSA data files. You can either:
+Download and setup M-CSA data automatically:
 
-**Option A: Download automatically (recommended)**
 ```python
-from csa3d import config
-config.setup_data("/path/to/csa3d_data")  # Downloads ~50 GB
+from csa3d import setup_data
+
+# Downloads ~50 GB of data to ~/.csa3d/data
+# Set up specific location with:
+# setup_data("/path/to/csa3d_data")
+success = setup_data()
+
+if success:
+    print("✓ Data setup complete!")
 ```
 
-**Option B: Set environment variable**
+Or via command line:
+```bash
+python -m csa3d.data_setup /path/to/csa3d_data
+```
+
+### 2. Manual Data Configuration
+
+If you prefer to download data manually:
+
+**Option A: Environment variable**
 ```bash
 export CSA3D_DATA_DIR="/path/to/csa3d_data"
 ```
 
-**Option C: Set in code**
+**Option B: Set in code**
 ```python
-from csa3d import config
-config.set_data_dir("/path/to/csa3d_data")
+from csa3d import set_data_dir
+set_data_dir("/path/to/csa3d_data")
 ```
 
-### 2. Load and Inspect Active Sites
+### 3. Load and Inspect Active Sites
 
 ```python
 from csa3d import Mcsa
@@ -151,7 +167,7 @@ for entry in db:
         print(f"    Resolution: {site.resolution} Å")
 ```
 
-### 3. Clustering and Template Creation
+### 4. Clustering and Template Creation
 
 ```python
 # Build multiple entries

{csa3d-0.2.0 → csa3d-0.2.2}/csa3d.egg-info/SOURCES.txt

@@ -1,11 +1,15 @@
 LICENSE
 README.md
 pyproject.toml
+csa3d/__init__.py
+csa3d/data_setup.py
 csa3d.egg-info/PKG-INFO
 csa3d.egg-info/SOURCES.txt
 csa3d.egg-info/dependency_links.txt
 csa3d.egg-info/requires.txt
 csa3d.egg-info/top_level.txt
+csa3d/_vendor/__init__.py
+csa3d/_vendor/dendrogram_cut.py
 csa3d/config/__init__.py
 csa3d/config/loader.py
 csa3d/config/settings.py

{csa3d-0.2.0 → csa3d-0.2.2}/csa3d.egg-info/requires.txt

@@ -5,10 +5,12 @@ scipy>=1.7.0
 scikit-learn>=0.24.0
 rdkit>=2021.09.0
 natsort>=8.0.0
-dendrogram-cut>=0.1.0
-parity-core>=0.5.0
 rmsd>=1.4.0
 
+[clustering]
+matplotlib>=3.5
+plotly>=5.0
+
 [dev]
 pytest>=7.0
 pytest-cov>=4.0

{csa3d-0.2.0 → csa3d-0.2.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "csa3d"
-version = "0.2.0"
+version = "0.2.2"
 description = "Extract and analyze enzyme catalytic sites in 3D from the Mechanism and Catalytic Site Atlas (M-CSA)"
 readme = "README.md"
 license = {text = "Apache-2.0"}
@@ -34,12 +34,14 @@ dependencies = [
     "scikit-learn>=0.24.0",
     "rdkit>=2021.09.0",
     "natsort>=8.0.0",
-    "dendrogram-cut>=0.1.0",
-    "parity-core>=0.5.0",
     "rmsd>=1.4.0",
 ]
 
 [project.optional-dependencies]
+clustering = [
+    "matplotlib>=3.5",
+    "plotly>=5.0",
+]
 dev = [
     "pytest>=7.0",
     "pytest-cov>=4.0",
@@ -68,7 +70,7 @@ Repository = "https://github.com/your-username/csa3d"
 "Issue Tracker" = "https://github.com/your-username/csa3d/issues"
 
 [tool.setuptools]
-packages = ["csa3d", "csa3d.config"]
+packages = ["csa3d", "csa3d.config", "csa3d._vendor"]
 
 [tool.black]
 line-length = 100