Refactored code

Author: satvshr

pyaptamer/AptaNet/pseaac.py

@@ -1,7 +1,7 @@
 import numpy as np
 from _props import aa_props
 
-from pyaptamer.utils import is_valid_aa
+from pyaptamer.utils import AMINO_ACIDS, is_valid_aa
 
 
 class PSeAAC:
@@ -32,28 +32,6 @@ def __init__(self):
         """
         Initialize PSeAAC with a protein sequence.
         """
-        self.amino_acid = [
-            "A",
-            "C",
-            "D",
-            "E",
-            "F",
-            "G",
-            "H",
-            "I",
-            "K",
-            "L",
-            "M",
-            "N",
-            "P",
-            "Q",
-            "R",
-            "S",
-            "T",
-            "V",
-            "W",
-            "Y",
-        ]
 
         # Load normalized property matrix (20x21, rows=AA, cols=NP1-NP21)
         self.np_matrix = aa_props(type="numpy", normalize=True)
@@ -86,10 +64,8 @@ def _average_aa(self, seq):
         from collections import Counter
 
         counts = Counter(seq)
-        total = len(self.amino_acid)
-        return {
-            aa: counts.get(aa, 0) / total if total > 0 else 0 for aa in self.amino_acid
-        }
+        total = len(AMINO_ACIDS)
+        return {aa: counts.get(aa, 0) / total if total > 0 else 0 for aa in AMINO_ACIDS}
 
     def _theta_rirj(self, ri, rj, prop_group):
         """
@@ -109,8 +85,8 @@ def _theta_rirj(self, ri, rj, prop_group):
         float
             Theta value.
         """
-        idx_ri = self.amino_acid.index(ri)
-        idx_rj = self.amino_acid.index(rj)
+        idx_ri = AMINO_ACIDS.index(ri)
+        idx_rj = AMINO_ACIDS.index(rj)
         diffs = (
             self.np_matrix[idx_rj, list(prop_group)]
             - self.np_matrix[idx_ri, list(prop_group)]
@@ -151,7 +127,7 @@ def transform(self, protein_sequence):
         if not is_valid_aa(protein_sequence):
             raise ValueError(
                 "Invalid amino acid found in protein_sequence. Only "
-                f"{''.join(self.amino_acid)} are allowed."
+                f"{''.join(AMINO_ACIDS)} are allowed."
             )
 
         lambda_val = 30
@@ -181,7 +157,7 @@ def transform(self, protein_sequence):
             denominator_val = sum_all_aa_freq + (weight * sum_all_theta_val)
 
             # First 20 features: normalized amino acid composition
-            aa_composition = np.array([aa_freq[aa] for aa in self.amino_acid])
+            aa_composition = np.array([aa_freq[aa] for aa in AMINO_ACIDS])
             all_pseaac.extend(np.round(aa_composition / denominator_val, 3))
 
             # Next 30 features: theta values

pyaptamer/AptaNet/test_pseaac.py

@@ -4,24 +4,6 @@
 from pseaac import PSeAAC
 
 
-def _normalize_properties(property_dicts):
-    """
-    Takes multiple amino acid property dictionaries and returns their
-    normalized versions.
-    Normalization: (value - mean) / std deviation
-    Returns a list of normalized dictionaries in the same order.
-    """
-    normalized = []
-    for prop in property_dicts:
-        values = list(prop.values())
-        mean_val = sum(values) / len(values)
-        std_val = (sum((v - mean_val) ** 2 for v in values) / len(values)) ** 0.5
-        normalized.append(
-            {aa: round((val - mean_val) / std_val, 3) for aa, val in prop.items()}
-        )
-    return normalized
-
-
 def test_normalized_values():
     """
     Test that normalized property matrix matches expected normalized values.