calibration active for webtool

tiger.py

@@ -178,8 +178,9 @@ def get_on_target_predictions(transcripts: pd.DataFrame, model: tf.keras.Model,
         target_seq, guide_seq, model_inputs = process_data(row[SEQ_COL])
 
         # get predictions
-        lfc_estimate = model.predict(model_inputs, batch_size=BATCH_SIZE_COMPUTE, verbose=False)
-        scores = transform_predictions(tf.squeeze(lfc_estimate).numpy())
+        lfc_estimate = np.squeeze(model.predict(model_inputs, batch_size=BATCH_SIZE_COMPUTE, verbose=False))
+        lfc_estimate = calibrate_predictions(lfc_estimate, num_mismatches=np.zeros_like(lfc_estimate))
+        scores = transform_predictions(lfc_estimate)
         predictions = pd.concat([predictions, pd.DataFrame({
             ID_COL: [index] * len(scores),
             TARGET_COL: target_seq,
@@ -306,8 +307,9 @@ def predict_off_target(off_targets: pd.DataFrame, model: tf.keras.Model):
         tf.reshape(one_hot_encode_sequence(off_targets[TARGET_COL], add_context_padding=False), [len(off_targets), -1]),
         tf.reshape(one_hot_encode_sequence(off_targets[GUIDE_COL], add_context_padding=True), [len(off_targets), -1]),
         ], axis=-1)
-    lfc_estimate = model.predict(model_inputs, batch_size=BATCH_SIZE_COMPUTE, verbose=False)
-    off_targets[SCORE_COL] = transform_predictions(tf.squeeze(lfc_estimate).numpy())
+    lfc_estimate = np.squeeze(model.predict(model_inputs, batch_size=BATCH_SIZE_COMPUTE, verbose=False))
+    lfc_estimate = calibrate_predictions(lfc_estimate, off_targets['Number of Mismatches'].to_numpy())
+    off_targets[SCORE_COL] = transform_predictions(lfc_estimate)
 
     return off_targets.reset_index(drop=True)