updates to the package synced with notebooks

Author: ramosv

bioneuralnet/datasets/kipan/clinical.csv

@@ -1,4 +1,4 @@
-Patient_ID,vital_status
+Patient_ID,target
 TCGA-CS-4938,0
 TCGA-CS-4941,1
 TCGA-CS-4942,1

bioneuralnet/datasets/lgg/clinical.csv

@@ -6,7 +6,7 @@
 from .correlation import (
     omics_correlation,
     cluster_correlation,
-    louvain_to_adjacency,
+    cluster_pca_correlation,
 )
 
 from .plot import (
@@ -23,7 +23,7 @@
 __all__ = [
     "omics_correlation",
     "cluster_correlation",
-    "louvain_to_adjacency",
+    "cluster_pca_correlation",
     "plot_variance_distribution",
     "plot_variance_by_feature",
     "plot_performance_three",

bioneuralnet/datasets/lgg/meth.csv

@@ -44,7 +44,7 @@ def omics_correlation(omics: pd.DataFrame, pheno: pd.DataFrame) -> Tuple[float,
     return corr, pvalue
 
 
-def cluster_correlation(cluster_df: pd.DataFrame, pheno: pd.DataFrame) -> tuple:
+def cluster_pca_correlation(cluster_df: pd.DataFrame, pheno: pd.DataFrame) -> tuple:
     """Computes the Pearson correlation coefficient between PC1 of a cluster and phenotype.
 
     Args:
@@ -88,12 +88,15 @@ def cluster_correlation(cluster_df: pd.DataFrame, pheno: pd.DataFrame) -> tuple:
 
     return (cluster_size, corr)
 
-def louvain_to_adjacency(louvain_cluster: pd.DataFrame) -> pd.DataFrame:
-    """Converts a Louvain cluster to an adjacency matrix.
+def cluster_correlation(louvain_cluster: pd.DataFrame) -> pd.DataFrame:
+    """Builds a new correlation network from an extracted cluster/subnetwork.
+    
+    This function is often used as an intermediate step between HybridLouvain and plotting. 
+    It allows the subnetwork to reveal its internal topological structure by computing the Pearson correlation of node connectivity profiles.
 
     Args:
 
-        louvain_cluster (pd.DataFrame): Represents an induced subnetwork (from Louvain).
+        louvain_cluster (pd.DataFrame): Represents an induced subnetwork.
 
     Returns:
 
@@ -103,4 +106,4 @@ def louvain_to_adjacency(louvain_cluster: pd.DataFrame) -> pd.DataFrame:
     np.fill_diagonal(adjacency_matrix.values, 0)
     adjacency_matrix = adjacency_matrix.fillna(0)
 
-    return adjacency_matrix
+    return adjacency_matrix

bioneuralnet/datasets/lgg/methylation.csv

@@ -7,7 +7,7 @@
 import matplotlib.cm as cm
 from sklearn.manifold import TSNE
 
-from .correlation import cluster_correlation
+from .correlation import cluster_pca_correlation
 from ..utils import get_logger
 
 logger = get_logger(__name__)
@@ -458,8 +458,8 @@ def get_size(pair):
     results = []
 
     for i, (df1, df2) in enumerate(zip(clusters1, clusters2), start=1):
-        size1, corr1 = cluster_correlation(df1, pheno)
-        size2, corr2 = cluster_correlation(df2, pheno)
+        size1, corr1 = cluster_pca_correlation(df1, pheno)
+        size2, corr2 = cluster_pca_correlation(df2, pheno)
 
         if corr1 is not None and corr2 is not None:
             results.append((f"Cluster_{i}", size1, corr1, size2, corr2))