Improve text in experiment tab and frame as ab discovery

folding_studio_demo/app.py

@@ -266,21 +266,28 @@ def model_comparison(api_key: str) -> None:
         )
 
 
-def create_correlation_tab():
-    gr.Markdown("# Correlation with experimental binding affinity data")
+def create_antibody_discovery_tab():
+    gr.Markdown("# Accelerating Antibody Discovery: In-Silico and Experimental Insights")
     gr.Markdown("""
-        This analysis explores the relationship between protein folding model confidence scores and experimental binding affinity data.
+        This analysis explores the relationship between protein folding model confidence scores and experimental binding affinity data, with the goal of accelerating antibody drug discovery.
         
-        The experimental dataset contains binding affinity measurements (KD in nM) between antibody-antigen pairs. 
+        The experimental dataset contains binding affinity measurements (KD in nM) between different antibodies and a fixed antigen target. 
         Each data point includes:
         - The antibody's light and heavy chain sequences
-        - The antigen sequence 
-        - The experimental KD value
+        - The antigen sequence
+        - The experimental KD value (binding strength)
         
         The analysis involves submitting these sequences to protein folding models for 3D structure prediction.
-        The models generate various confidence scores for each prediction. These scores are then correlated
-        with the experimental binding affinity measurements to evaluate their effectiveness as predictors
-        of binding strength.
+        The models generate various confidence scores for each prediction. By correlating these scores
+        with the experimental binding affinity measurements, we can identify which confidence metrics
+        are the best predictors of actual binding strength.
+        
+        This correlation study has important implications for drug discovery: once we identify a reliable
+        computational predictor of binding affinity, we can use it to rapidly screen thousands of candidate
+        antibody sequences in-silico. This computational approach is much faster than experimental testing,
+        allowing us to efficiently identify promising antibody candidates that are likely to bind strongly
+        to the target antigen. The best candidates can then be validated experimentally, significantly
+        accelerating the drug discovery process.
     """)
     spr_data_with_scores = pd.read_csv("spr_af_scores_mapped.csv")
     spr_data_with_scores = spr_data_with_scores.rename(columns=SCORE_COLUMN_NAMES)
@@ -429,7 +436,7 @@ def __main__():
             simple_prediction(api_key)
         with gr.Tab("📊 Model Comparison"):
             model_comparison(api_key)
-        with gr.Tab("🔍 Correlations"):
-            create_correlation_tab()
+        with gr.Tab("🧪 Antibody Discovery"):
+            create_antibody_discovery_tab()
 
     demo.launch()

folding_studio_demo/correlate.py

@@ -4,7 +4,7 @@ from pathlib import Path
 import numpy as np
 import pandas as pd
 import plotly.graph_objects as go
-from scipy.stats import linregress, pearsonr, spearmanr
+from scipy.stats import pearsonr, spearmanr
 
 logger = logging.getLogger(__name__)