Update app.py
Browse files
app.py
CHANGED
@@ -14,39 +14,9 @@ import urllib.parse
|
|
14 |
|
15 |
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
|
16 |
|
17 |
-
# Set page config and apply dark theme
|
18 |
st.set_page_config(layout='wide')
|
19 |
-
st.
|
20 |
-
|
21 |
-
body {
|
22 |
-
color: #fff;
|
23 |
-
background-color: #0e1117;
|
24 |
-
}
|
25 |
-
.stApp {
|
26 |
-
background-color: #0e1117;
|
27 |
-
}
|
28 |
-
.stTextInput > div > div > input {
|
29 |
-
color: #fff;
|
30 |
-
background-color: #262730;
|
31 |
-
}
|
32 |
-
.stNumberInput > div > div > input {
|
33 |
-
color: #fff;
|
34 |
-
background-color: #262730;
|
35 |
-
}
|
36 |
-
.stTextArea > div > div > textarea {
|
37 |
-
color: #fff;
|
38 |
-
background-color: #262730;
|
39 |
-
}
|
40 |
-
.stButton > button {
|
41 |
-
color: #fff;
|
42 |
-
background-color: #0e1117;
|
43 |
-
border: 1px solid #fff;
|
44 |
-
}
|
45 |
-
</style>
|
46 |
-
""", unsafe_allow_html=True)
|
47 |
-
|
48 |
-
st.title('🔮 GenPro2 Protein Generator, Structure Predictor, and Analysis Tool')
|
49 |
-
st.write('GenPro2 is an end-to-end protein sequence generator, structure predictor, and analysis tool based [*ESMFold*](https://esmatlas.com/about) and the ESM-2 language model.')
|
50 |
|
51 |
def generate_sequence_from_words(words, length):
|
52 |
seed = ' '.join(words).encode('utf-8')
|
@@ -94,10 +64,11 @@ def perform_blast_analysis(sequence):
|
|
94 |
st.write(f"**Organism Code:** {organism}")
|
95 |
st.write(f"**Sequence Identity:** {identity_percentage:.2f}%")
|
96 |
|
|
|
97 |
# Fetch protein function (if available)
|
98 |
if hasattr(alignment, 'description') and alignment.description:
|
99 |
st.write(f"**Potential Function:** {alignment.description}")
|
100 |
-
|
101 |
st.write("No significant matches found. This might be a unique protein sequence!")
|
102 |
except Exception as e:
|
103 |
st.error(f"An error occurred during protein analysis: {str(e)}")
|
@@ -150,29 +121,25 @@ if 'show_analyze_button' not in st.session_state:
|
|
150 |
if 'structure_info' not in st.session_state:
|
151 |
st.session_state.structure_info = None
|
152 |
|
153 |
-
|
154 |
-
st.subheader("Generate Sequence from Words")
|
155 |
-
col1, col2, col3 = st.columns(3)
|
156 |
-
with col1:
|
157 |
-
word1 = st.text_input("Word 1")
|
158 |
-
with col2:
|
159 |
-
word2 = st.text_input("Word 2")
|
160 |
-
with col3:
|
161 |
-
word3 = st.text_input("Word 3")
|
162 |
|
163 |
-
|
|
|
|
|
|
|
|
|
164 |
|
165 |
-
if st.button('Generate and Predict'):
|
166 |
if word1 and word2 and word3:
|
167 |
sequence = generate_sequence_from_words([word1, word2, word3], sequence_length)
|
168 |
st.session_state.sequence = sequence
|
169 |
-
st.text_area("Generated Sequence", sequence, height=100)
|
170 |
-
st.info("Note: The same words and sequence length will always produce the same sequence.")
|
171 |
|
172 |
with st.spinner("Predicting protein structure... This may take a few minutes."):
|
173 |
update(sequence, word1, word2, word3, sequence_length)
|
174 |
else:
|
175 |
-
st.warning("Please enter all three words to generate a sequence.")
|
176 |
|
177 |
# Display structure information if available
|
178 |
if st.session_state.structure_info:
|
@@ -181,14 +148,14 @@ if st.session_state.structure_info:
|
|
181 |
render_mol(info['pdb_string'])
|
182 |
|
183 |
st.subheader('plDDT Confidence Score')
|
184 |
-
st.write('plDDT is a
|
185 |
plddt_score = int(info["b_value"] * 100)
|
186 |
st.info(f'Your plDDT score is: {plddt_score}%')
|
187 |
|
188 |
st.subheader("Share your unique protein on X(Twitter)")
|
189 |
|
190 |
st.markdown("""
|
191 |
-
<div style='background-color: #
|
192 |
<ol>
|
193 |
<li>Take a screenshot of the protein structure above.</li>
|
194 |
<li>Click the 'Share on X' button below to open a pre-filled post with your protein seed-words and score.</li>
|
|
|
14 |
|
15 |
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
|
16 |
|
|
|
17 |
st.set_page_config(layout='wide')
|
18 |
+
st.sidebar.title('🔮 GenPro2 Protein Generator, Structure Predictor, and Analysis Tool')
|
19 |
+
st.sidebar.write('GenPro2 is an end-to-end protein sequence generator, structure predictor, and analysis tool based [*ESMFold*](https://esmatlas.com/about) and the ESM-2 language model.')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
|
21 |
def generate_sequence_from_words(words, length):
|
22 |
seed = ' '.join(words).encode('utf-8')
|
|
|
64 |
st.write(f"**Organism Code:** {organism}")
|
65 |
st.write(f"**Sequence Identity:** {identity_percentage:.2f}%")
|
66 |
|
67 |
+
|
68 |
# Fetch protein function (if available)
|
69 |
if hasattr(alignment, 'description') and alignment.description:
|
70 |
st.write(f"**Potential Function:** {alignment.description}")
|
71 |
+
|
72 |
st.write("No significant matches found. This might be a unique protein sequence!")
|
73 |
except Exception as e:
|
74 |
st.error(f"An error occurred during protein analysis: {str(e)}")
|
|
|
121 |
if 'structure_info' not in st.session_state:
|
122 |
st.session_state.structure_info = None
|
123 |
|
124 |
+
st.title("Word-Seeded Protein Sequence Generator, Structure Predictor, and Analysis Tool")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
125 |
|
126 |
+
st.sidebar.subheader("Generate Sequence from Words")
|
127 |
+
word1 = st.sidebar.text_input("Word 1")
|
128 |
+
word2 = st.sidebar.text_input("Word 2")
|
129 |
+
word3 = st.sidebar.text_input("Word 3")
|
130 |
+
sequence_length = st.sidebar.number_input("Sequence Length", min_value=50, max_value=400, value=100, step=10)
|
131 |
|
132 |
+
if st.sidebar.button('Generate and Predict'):
|
133 |
if word1 and word2 and word3:
|
134 |
sequence = generate_sequence_from_words([word1, word2, word3], sequence_length)
|
135 |
st.session_state.sequence = sequence
|
136 |
+
st.sidebar.text_area("Generated Sequence", sequence, height=100)
|
137 |
+
st.sidebar.info("Note: The same words and sequence length will always produce the same sequence.")
|
138 |
|
139 |
with st.spinner("Predicting protein structure... This may take a few minutes."):
|
140 |
update(sequence, word1, word2, word3, sequence_length)
|
141 |
else:
|
142 |
+
st.sidebar.warning("Please enter all three words to generate a sequence.")
|
143 |
|
144 |
# Display structure information if available
|
145 |
if st.session_state.structure_info:
|
|
|
148 |
render_mol(info['pdb_string'])
|
149 |
|
150 |
st.subheader('plDDT Confidence Score')
|
151 |
+
st.write('plDDT is a bench mark for scoring the confidence level in protein folding predictions based on a scale from 0-100%. 70% or more is good!')
|
152 |
plddt_score = int(info["b_value"] * 100)
|
153 |
st.info(f'Your plDDT score is: {plddt_score}%')
|
154 |
|
155 |
st.subheader("Share your unique protein on X(Twitter)")
|
156 |
|
157 |
st.markdown("""
|
158 |
+
<div style='background-color: #e6f2ff; padding: 10px; border-radius: 5px; font-size: 0.8em;'>
|
159 |
<ol>
|
160 |
<li>Take a screenshot of the protein structure above.</li>
|
161 |
<li>Click the 'Share on X' button below to open a pre-filled post with your protein seed-words and score.</li>
|