removing statsmodels

ohsu-comp-bio · Feb 22, 2023 · bcf555d · bcf555d
1 parent 5ed01ac
commit bcf555d
Show file tree

Hide file tree

Showing 4 changed files with 1 addition and 9 deletions.
diff --git a/data/Homo_sapiens.GRCh38.95.gtf.gz b/data/Homo_sapiens.GRCh38.95.gtf.gz
diff --git a/data/Mus_musculus.GRCm38.95.gtf.gz b/data/Mus_musculus.GRCm38.95.gtf.gz
diff --git a/data/secondary_intx_regulon.pkl b/data/secondary_intx_regulon.pkl
diff --git a/priori/priori.py b/priori/priori.py
@@ -3,7 +3,6 @@
 import functools
 import pandas as pd
 import numpy as np
-import statsmodels.api as sm
 from tqdm import tqdm
 from sklearn.utils.validation import check_array
 import priori.regulon.regulon_enrichment as regulon_enrichment
@@ -76,7 +75,6 @@ def __init__(self, expr, regulon=None, regulon_size=15,
 
         self.correlations = None
         self.p_values = None
-        self.adjusted_p_values = None
 
     def __str__(self):
         return """------\nn-features: {}\nn-samples: {}\nscaler: {}\nscaled:\
@@ -259,15 +257,10 @@ def assign_weights(self):
 
         # Calculate spearman correlation
         r, p = regulon_utils.spearmanr(self.expr)
-
-        # # FDR-adjust p-values using the Benjamini-Hochberg method
-        bool, p_adj, sidak, bonf = sm.stats.multipletests(p.flatten(), method = "fdr_bh")
-        p_adj = p_adj.reshape(p.shape)
 
         # Create data frames
         r_frame = pd.DataFrame(r, columns=self.expr.columns, index=self.expr.columns)
         p_frame = pd.DataFrame(p, columns=self.expr.columns, index=self.expr.columns)
-        p_adj_frame = pd.DataFrame(p_adj, columns=self.expr.columns, index=self.expr.columns)
 
         # Calculate F regression to correct for collinearity
         F_statistics = {regulator: regulon_utils.f_regression(
@@ -280,13 +273,12 @@ def assign_weights(self):
                                                      pruned_regulon,
                                                      F_statistics,
                                                      r_frame,
-                                                     p_adj_frame)
+                                                     p_frame)
                              for regulator in F_statistics])
 
         self.regulon_weights = weights[~np.isinf(weights.MoA)]
         self.correlations = r_frame
         self.p_values = p_frame
-        self.adjusted_p_values = p_adj_frame
 
     def calculate_enrichment(self):
         """