Added model

amino_acid_composition.py

@@ -0,0 +1,45 @@
+# -*- coding: utf-8 -*-
+"""amino_acid_composition.ipynb
+
+Automatically generated by Colab.
+
+Original file is located at
+    https://colab.research.google.com/drive/1YebtHJU3a9oNapMztiEku0M2VToI_1Lm
+"""
+
+# amino_acid_composition.py
+
+def amino_acid_composition(sequence):
+    amino_acids = 'ACDEFGHIKLMNPQRSTVWY'
+    composition = {aa: 0 for aa in amino_acids}
+    total = len(sequence)
+
+    for aa in sequence:
+        if aa in composition:
+            composition[aa] += 1
+
+    for aa in composition:
+        composition[aa] = (composition[aa] / total) * 100
+
+    return composition
+
+def process_dataset(dataset):
+    compositions = []
+    for sequence in dataset:
+        compositions.append(amino_acid_composition(sequence))
+    return compositions
+
+def main(active_peptides, inactive_peptides):
+    active_compositions = process_dataset(active_peptides)
+    inactive_compositions = process_dataset(inactive_peptides)
+    return active_compositions, inactive_compositions
+
+if __name__ == "__main__":
+    # Example usage
+    active_peptides = ["ACDEFGHIKLMNPQRSTVWY", "ACDEFGHIKLMN"]
+    inactive_peptides = ["QRSTVWYACDEFGHIKLMN", "HIKLMNPQRST"]
+
+    active_compositions, inactive_compositions = main(active_peptides, inactive_peptides)
+    print("Active Peptide Compositions:", active_compositions)
+    print("Inactive Peptide Compositions:", inactive_compositions)
+

app.py

@@ -0,0 +1,224 @@
+# import streamlit as st
+# import pandas as pd
+# import joblib
+# from sklearn.ensemble import RandomForestClassifier
+# import matplotlib.pyplot as plt
+# import seaborn as sns
+
+# # Load the trained model (ensure the model file is in the same directory)
+# model = joblib.load('model.pkl')
+
+# # Function to process new peptide sequences
+# def process_peptide_sequences(peptides):
+#     # Example processing function, replace with actual preprocessing steps
+#     compositions = []
+#     for peptide in peptides:
+#         composition = {aa: peptide.count(aa) for aa in 'ACDEFGHIKLMNPQRSTVWY'}
+#         compositions.append(composition)
+#     return pd.DataFrame(compositions)
+
+# # Streamlit app
+# st.title("ABPep-C")
+# st.write("Classify peptide sequences as active or inactive against biofilm")
+
+# # Input: Peptide sequences
+# peptide_input = st.text_area("Enter peptide sequences (one per line)")
+# peptides = peptide_input.split('\n')
+
+# if st.button("Classify"):
+#     if peptides:
+#         # Process the input peptides
+#         peptide_df = process_peptide_sequences(peptides)
+        
+#         # Predict using the trained model
+#         predictions = model.predict(peptide_df)
+#         results = pd.DataFrame({
+#             'Peptide': peptides,
+#             'Prediction': predictions
+#         })
+#         results['Prediction'] = results['Prediction'].map({0: 'Inactive', 1: 'Active'})
+        
+#         # Display the results
+#         st.write("Classification Results")
+#         st.write(results)
+        
+#         # Display interactive graphs
+#         st.write("Prediction Distribution")
+#         fig, ax = plt.subplots()
+#         sns.countplot(x='Prediction', data=results, ax=ax)
+#         st.pyplot(fig)
+        
+#         st.write("Amino Acid Composition of Peptides")
+#         amino_acid_counts = peptide_df.sum().reset_index()
+#         amino_acid_counts.columns = ['Amino Acid', 'Count']
+#         fig, ax = plt.subplots()
+#         sns.barplot(x='Amino Acid', y='Count', data=amino_acid_counts, ax=ax)
+#         st.pyplot(fig)
+#     else:
+#         st.write("Please enter peptide sequences.")
+
+# # Save this script as app.py and run it using: streamlit run app.
+#######################################################################################################################################
+# import streamlit as st
+# import pandas as pd
+# import joblib
+# from sklearn.ensemble import RandomForestClassifier
+# import matplotlib.pyplot as plt
+# import seaborn as sns
+
+# # Load the trained model (ensure the model file is in the same directory)
+# model = joblib.load('model.pkl')
+
+# # Function to process new peptide sequences
+# def process_peptide_sequences(peptides):
+#     # Example processing function, replace with actual preprocessing steps
+#     compositions = []
+#     for peptide in peptides:
+#         composition = {aa: peptide.count(aa) for aa in 'ACDEFGHIKLMNPQRSTVWY'}
+#         compositions.append(composition)
+#     return pd.DataFrame(compositions)
+
+# # Custom CSS for font size and color
+# st.markdown("""
+#     <style>
+#     .title {
+#         font-size: 48px !important;
+#         color: #4CAF50;
+#     }
+#     .subheader {
+#         font-size: 24px !important;
+#         color: #FF5722;
+#     }
+#     .text {
+#         font-size: 18px !important;
+#     }
+#     </style>
+#     """, unsafe_allow_html=True)
+
+# # Streamlit app
+# st.markdown('<h1 class="title">Ab-PepC</h1>', unsafe_allow_html=True)
+# st.markdown('<h2 class="subheader">Classify peptide sequences as active or inactive against biofilm</h2>', unsafe_allow_html=True)
+
+# # Input: Peptide sequences
+# peptide_input = st.text_area("Enter peptide sequences (one per line)")
+# peptides = peptide_input.split('\n')
+
+# if st.button("Classify"):
+#     if peptides:
+#         # Process the input peptides
+#         peptide_df = process_peptide_sequences(peptides)
+        
+#         # Predict using the trained model
+#         predictions = model.predict(peptide_df)
+#         results = pd.DataFrame({
+#             'Peptide': peptides,
+#             'Prediction': predictions
+#         })
+#         results['Prediction'] = results['Prediction'].map({0: 'Inactive', 1: 'Active'})
+        
+#         # Display the results
+#         st.markdown('<h3 class="subheader">Classification Results</h3>', unsafe_allow_html=True)
+#         st.dataframe(results)
+        
+#         # Display interactive graphs
+#         st.markdown('<h3 class="subheader">Prediction Distribution</h3>', unsafe_allow_html=True)
+#         fig, ax = plt.subplots()
+#         sns.countplot(x='Prediction', data=results, ax=ax)
+#         ax.set_xlabel('Prediction', fontsize=18)
+#         ax.set_ylabel('Count', fontsize=18)
+#         st.pyplot(fig)
+        
+#         st.markdown('<h3 class="subheader">Amino Acid Composition of Peptides</h3>', unsafe_allow_html=True)
+#         amino_acid_counts = peptide_df.sum().reset_index()
+#         amino_acid_counts.columns = ['Amino Acid', 'Count']
+#         fig, ax = plt.subplots()
+#         sns.barplot(x='Amino Acid', y='Count', data=amino_acid_counts, ax=ax)
+#         ax.set_xlabel('Amino Acid', fontsize=18)
+#         ax.set_ylabel('Count', fontsize=18)
+#         st.pyplot(fig)
+#     else:
+#         st.write("Please enter peptide sequences.")
+#######################################################################################################################################
+import streamlit as st
+import pandas as pd
+import joblib
+from sklearn.ensemble import RandomForestClassifier
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+# Load the trained model (ensure the model file is in the same directory)
+model = joblib.load('model.pkl')
+
+# Function to process new peptide sequences
+def process_peptide_sequences(peptides):
+    # Example processing function, replace with actual preprocessing steps
+    compositions = []
+    for peptide in peptides:
+        composition = {aa: peptide.count(aa) for aa in 'ACDEFGHIKLMNPQRSTVWY'}
+        compositions.append(composition)
+    return pd.DataFrame(compositions)
+
+# Custom CSS for font size and color
+st.markdown("""
+    <style>
+    .title {
+        font-size: 48px !important;
+        color: #4CAF50;
+    }
+    .subheader {
+        font-size: 24px !important;
+        color: #FF5722;
+    }
+    .text {
+        font-size: 18px !important;
+    }
+    </style>
+    """, unsafe_allow_html=True)
+
+# Streamlit app
+col1, col2 = st.columns([1, 4])  # Adjust the width ratio as needed
+col1.image('Ab-PepC_logo.png', width=150)  # Add your logo file path here
+with col2:
+    st.markdown('<h1 class="title">ABPep-C</h1>', unsafe_allow_html=True)
+    st.markdown('<h2 class="subheader">Classify peptide sequences as active or inactive against biofilm</h2>', unsafe_allow_html=True)
+
+# Input: Peptide sequences
+peptide_input = st.text_area("Enter peptide sequences (one per line)")
+peptides = peptide_input.split('\n')
+
+if st.button("Classify"):
+    if peptides:
+        # Process the input peptides
+        peptide_df = process_peptide_sequences(peptides)
+        
+        # Predict using the trained model
+        predictions = model.predict(peptide_df)
+        results = pd.DataFrame({
+            'Peptide': peptides,
+            'Prediction': predictions
+        })
+        results['Prediction'] = results['Prediction'].map({0: 'Inactive', 1: 'Active'})
+        
+        # Display the results
+        st.markdown('<h3 class="subheader">Classification Results</h3>', unsafe_allow_html=True)
+        st.dataframe(results)
+        
+        # Display interactive graphs
+        st.markdown('<h3 class="subheader">Prediction Distribution</h3>', unsafe_allow_html=True)
+        fig, ax = plt.subplots()
+        sns.countplot(x='Prediction', data=results, ax=ax)
+        ax.set_xlabel('Prediction', fontsize=18)
+        ax.set_ylabel('Count', fontsize=18)
+        st.pyplot(fig)
+        
+        st.markdown('<h3 class="subheader">Amino Acid Composition of Peptides</h3>', unsafe_allow_html=True)
+        amino_acid_counts = peptide_df.sum().reset_index()
+        amino_acid_counts.columns = ['Amino Acid', 'Count']
+        fig, ax = plt.subplots()
+        sns.barplot(x='Amino Acid', y='Count', data=amino_acid_counts, ax=ax)
+        ax.set_xlabel('Amino Acid', fontsize=18)
+        ax.set_ylabel('Count', fontsize=18)
+        st.pyplot(fig)
+    else:
+        st.write("Please enter peptide sequences.")
+

model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9458d55fb443d797241263126170928459672aad1a93864939bbe36092c8394d
+size 309433

requirements.txt

@@ -0,0 +1,6 @@
+streamlit
+pandas
+joblib
+scikit-learn
+matplotlib
+seaborn