Update app.py

app.py

@@ -6,6 +6,7 @@ import torch
 import torch.nn as nn
 import torch.nn.functional as F
 import xgboost as xgb
+import scipy.stats as stats
 import plotly.express as px
 import plotly.graph_objects as go
 from rdkit import Chem, DataStructs
@@ -136,19 +137,38 @@ def generate_and_screen(num_to_gen):
         
         mol = Chem.MolFromSmiles(generated.replace('*', 'C'))
         if mol:
+            # Diversity Check
             fp = rdFingerprintGenerator.GetMorganGenerator(radius=2).GetFingerprint(mol)
             if any(DataStructs.TanimotoSimilarity(fp, prev_fp) > 0.7 for prev_fp in generated_fps): continue
             
             feats = get_mff_features(generated)
-            s = np.expm1(strength_model.predict(feats.reshape(1, -1))[0])
+            s = np.expm1(strength_model.predict(feats.reshape(1, -1)))[0]
             d = degrad_model.predict(feats.reshape(1, -1))[0]
             t = tg_model.predict(feats.reshape(1, -1))[0]
             sa = sascorer.calculateScore(mol) if sascorer else 5.0
 
-            s_err, t_err, d_err = s * 0.05, 10.5, 0.05
-            passes = sum([(s-s_err) > T_STRENGTH, (t-t_err) > T_TG, (d-d_err) > T_DEG, sa < T_SA])
-            conf_pct = (passes / 4) * 100
-            status = "Excellent" if passes >= 3 else "Review" if passes == 2 else "Bad"
+            # --- DYNAMIC UNCERTAINTY (Points 2 & 4) ---
+            mol_wt = Descriptors.MolWt(mol)
+            noise = (mol_wt / 500.0) 
+            s_err = s * (0.04 + (0.03 * noise)) 
+            t_err = 8.0 + (5.0 * noise)
+            d_err = 0.03 + (0.04 * noise)
+
+            # --- CONTINUOUS CONFIDENCE (Normal Distribution / Z-score) ---
+            def calc_prob(val, target, err, higher_is_better=True):
+                z = (val - target) / (err + 1e-6)
+                prob = stats.norm.cdf(z)
+                return prob if higher_is_better else (1 - prob)
+
+            p_s = calc_prob(s, T_STRENGTH, s_err)
+            p_t = calc_prob(t, T_TG, t_err)
+            p_d = calc_prob(d, T_DEG, d_err)
+            p_sa = calc_prob(sa, T_SA, 1.0, False)
+
+            conf_pct = ((p_s + p_t + p_d + p_sa) / 4) * 100
+            if conf_pct >= 80: status = "Excellent"
+            elif conf_pct >= 55: status = "Review"
+            else: status = "Bad"
             
             generated_fps.append(fp)
             candidates.append({
@@ -164,12 +184,11 @@ def generate_and_screen(num_to_gen):
             
     df = pd.DataFrame(candidates)
     
-    # 1. FIXED COLORS & PARETO LINE
+    # Pareto Front Calculation & Fix
     df = df.sort_values(by="_s")
-    # Simple Pareto calculation (running max of Y for sorted X)
     pareto_x, pareto_y = [], []
     current_max_y = -1
-    for i, row in df.sort_values(by="_s").iterrows():
+    for i, row in df.iterrows():
         if row["_t"] > current_max_y:
             pareto_x.append(row["_s"])
             pareto_y.append(row["_t"])