upload files
Browse files- Dockerfile +3 -2
- app.py +386 -142
- requirements.txt +7 -4
Dockerfile
CHANGED
|
@@ -8,6 +8,7 @@ RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
|
|
| 8 |
|
| 9 |
COPY . .
|
| 10 |
|
| 11 |
-
EXPOSE
|
|
|
|
|
|
|
| 12 |
|
| 13 |
-
CMD ["shiny", "run", "app.py", "--host", "0.0.0.0", "--port", "7860"]
|
|
|
|
| 8 |
|
| 9 |
COPY . .
|
| 10 |
|
| 11 |
+
EXPOSE 8000
|
| 12 |
+
|
| 13 |
+
CMD ["python", "app.py"]
|
| 14 |
|
|
|
app.py
CHANGED
|
@@ -1,162 +1,406 @@
|
|
| 1 |
-
|
| 2 |
-
import
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
|
| 4 |
-
# Load
|
| 5 |
-
|
| 6 |
-
from shinywidgets import render_plotly
|
| 7 |
|
| 8 |
-
|
| 9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
|
| 11 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
|
| 13 |
-
|
| 14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
max=bill_rng[1],
|
| 22 |
-
value=bill_rng,
|
| 23 |
-
pre="$",
|
| 24 |
-
)
|
| 25 |
-
ui.input_checkbox_group(
|
| 26 |
-
"time",
|
| 27 |
-
"Food service",
|
| 28 |
-
["Lunch", "Dinner"],
|
| 29 |
-
selected=["Lunch", "Dinner"],
|
| 30 |
-
inline=True,
|
| 31 |
-
)
|
| 32 |
-
ui.input_action_button("reset", "Reset filter")
|
| 33 |
-
|
| 34 |
-
# Add main content
|
| 35 |
-
ICONS = {
|
| 36 |
-
"user": fa.icon_svg("user", "regular"),
|
| 37 |
-
"wallet": fa.icon_svg("wallet"),
|
| 38 |
-
"currency-dollar": fa.icon_svg("dollar-sign"),
|
| 39 |
-
"ellipsis": fa.icon_svg("ellipsis"),
|
| 40 |
-
}
|
| 41 |
-
|
| 42 |
-
with ui.layout_columns(fill=False):
|
| 43 |
-
with ui.value_box(showcase=ICONS["user"]):
|
| 44 |
-
"Total tippers"
|
| 45 |
-
|
| 46 |
-
@render.express
|
| 47 |
-
def total_tippers():
|
| 48 |
-
tips_data().shape[0]
|
| 49 |
-
|
| 50 |
-
with ui.value_box(showcase=ICONS["wallet"]):
|
| 51 |
-
"Average tip"
|
| 52 |
-
|
| 53 |
-
@render.express
|
| 54 |
-
def average_tip():
|
| 55 |
-
d = tips_data()
|
| 56 |
-
if d.shape[0] > 0:
|
| 57 |
-
perc = d.tip / d.total_bill
|
| 58 |
-
f"{perc.mean():.1%}"
|
| 59 |
-
|
| 60 |
-
with ui.value_box(showcase=ICONS["currency-dollar"]):
|
| 61 |
-
"Average bill"
|
| 62 |
-
|
| 63 |
-
@render.express
|
| 64 |
-
def average_bill():
|
| 65 |
-
d = tips_data()
|
| 66 |
-
if d.shape[0] > 0:
|
| 67 |
-
bill = d.total_bill.mean()
|
| 68 |
-
f"${bill:.2f}"
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
with ui.layout_columns(col_widths=[6, 6, 12]):
|
| 72 |
-
with ui.card(full_screen=True):
|
| 73 |
-
ui.card_header("Tips data")
|
| 74 |
-
|
| 75 |
-
@render.data_frame
|
| 76 |
-
def table():
|
| 77 |
-
return render.DataGrid(tips_data())
|
| 78 |
-
|
| 79 |
-
with ui.card(full_screen=True):
|
| 80 |
-
with ui.card_header(class_="d-flex justify-content-between align-items-center"):
|
| 81 |
-
"Total bill vs tip"
|
| 82 |
-
with ui.popover(title="Add a color variable", placement="top"):
|
| 83 |
-
ICONS["ellipsis"]
|
| 84 |
-
ui.input_radio_buttons(
|
| 85 |
-
"scatter_color",
|
| 86 |
-
None,
|
| 87 |
-
["none", "sex", "smoker", "day", "time"],
|
| 88 |
-
inline=True,
|
| 89 |
-
)
|
| 90 |
|
| 91 |
-
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
return px.scatter(
|
| 95 |
-
tips_data(),
|
| 96 |
-
x="total_bill",
|
| 97 |
-
y="tip",
|
| 98 |
-
color=None if color == "none" else color,
|
| 99 |
-
trendline="lowess",
|
| 100 |
-
)
|
| 101 |
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 113 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 114 |
|
| 115 |
-
|
| 116 |
-
def tip_perc():
|
| 117 |
-
from ridgeplot import ridgeplot
|
| 118 |
-
|
| 119 |
-
dat = tips_data()
|
| 120 |
-
dat["percent"] = dat.tip / dat.total_bill
|
| 121 |
-
yvar = input.tip_perc_y()
|
| 122 |
-
uvals = dat[yvar].unique()
|
| 123 |
|
| 124 |
-
|
| 125 |
|
| 126 |
-
|
| 127 |
-
|
| 128 |
-
|
| 129 |
-
|
| 130 |
-
colorscale="viridis",
|
| 131 |
-
colormode="row-index",
|
| 132 |
-
)
|
| 133 |
|
| 134 |
-
|
| 135 |
-
legend=dict(
|
| 136 |
-
orientation="h", yanchor="bottom", y=1.02, xanchor="center", x=0.5
|
| 137 |
-
)
|
| 138 |
-
)
|
| 139 |
|
| 140 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 141 |
|
|
|
|
| 142 |
|
| 143 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 144 |
|
| 145 |
-
#
|
| 146 |
-
|
| 147 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 148 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 149 |
|
| 150 |
-
|
| 151 |
-
|
| 152 |
-
|
| 153 |
-
|
| 154 |
-
|
| 155 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 156 |
|
|
|
|
| 157 |
|
| 158 |
-
|
| 159 |
-
|
| 160 |
-
|
| 161 |
-
|
| 162 |
-
ui.update_checkbox_group("time", selected=["Lunch", "Dinner"])
|
|
|
|
| 1 |
+
from transformers import pipeline
|
| 2 |
+
from rcsbsearchapi import TextQuery, AttributeQuery, Query
|
| 3 |
+
from rcsbsearchapi.search import Sort, SequenceQuery
|
| 4 |
+
import os
|
| 5 |
+
from dotenv import load_dotenv
|
| 6 |
+
from shiny import App, render, ui, reactive
|
| 7 |
+
import pandas as pd
|
| 8 |
+
import warnings
|
| 9 |
+
import re
|
| 10 |
+
from UniprotKB_P_Sequence_RCSB_API_test import ProteinQuery, ProteinSearchEngine
|
| 11 |
+
import plotly.graph_objects as go
|
| 12 |
+
from shinywidgets import output_widget, render_widget
|
| 13 |
+
warnings.filterwarnings('ignore')
|
| 14 |
|
| 15 |
+
# Load environment variables from .env file
|
| 16 |
+
load_dotenv()
|
|
|
|
| 17 |
|
| 18 |
+
class PDBSearchAssistant:
|
| 19 |
+
def __init__(self, model_name="google/flan-t5-large"):
|
| 20 |
+
# Set up HuggingFace pipeline with better model
|
| 21 |
+
self.pipe = pipeline(
|
| 22 |
+
"text2text-generation",
|
| 23 |
+
model=model_name,
|
| 24 |
+
max_new_tokens=512,
|
| 25 |
+
temperature=0.3,
|
| 26 |
+
torch_dtype="auto",
|
| 27 |
+
device="cpu"
|
| 28 |
+
)
|
| 29 |
+
|
| 30 |
+
self.prompt_template = """
|
| 31 |
+
Extract specific search parameters from the query, if present:
|
| 32 |
+
1. Resolution cutoff (in Γ
)
|
| 33 |
+
2. Sequence information
|
| 34 |
+
3. Specific PDB ID
|
| 35 |
+
4. Experimental method (X-RAY, EM, NMR)
|
| 36 |
|
| 37 |
+
Format:
|
| 38 |
+
Resolution: [maximum resolution in Γ
, if mentioned]
|
| 39 |
+
Sequence: [any sequence mentioned]
|
| 40 |
+
PDB_ID: [specific PDB ID if mentioned]
|
| 41 |
+
Method: [experimental method if mentioned]
|
| 42 |
|
| 43 |
+
Examples:
|
| 44 |
+
Query: "Find X-ray structures better than 2.5Γ
resolution"
|
| 45 |
+
Resolution: 2.5
|
| 46 |
+
Sequence: none
|
| 47 |
+
PDB_ID: none
|
| 48 |
+
Method: X-RAY
|
| 49 |
|
| 50 |
+
Query: "Show me NMR structures of kinases"
|
| 51 |
+
Resolution: none
|
| 52 |
+
Sequence: none
|
| 53 |
+
PDB_ID: none
|
| 54 |
+
Method: NMR
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
|
| 56 |
+
Now analyze:
|
| 57 |
+
Query: {query}
|
| 58 |
+
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
+
def search_pdb(self, query):
|
| 61 |
+
try:
|
| 62 |
+
# Get search parameters from LLM
|
| 63 |
+
formatted_prompt = self.prompt_template.format(query=query)
|
| 64 |
+
response = self.pipe(formatted_prompt)[0]['generated_text']
|
| 65 |
+
print("Generated parameters:", response)
|
| 66 |
+
|
| 67 |
+
# Parse LLM response
|
| 68 |
+
resolution_limit = None
|
| 69 |
+
pdb_id = None
|
| 70 |
+
sequence = None
|
| 71 |
+
method = None
|
| 72 |
+
has_resolution_query = False
|
| 73 |
+
resolution_direction = "less"
|
| 74 |
+
|
| 75 |
+
# Check if query contains resolution-related terms
|
| 76 |
+
resolution_terms = {
|
| 77 |
+
'better': 'less',
|
| 78 |
+
'best': 'less',
|
| 79 |
+
'highest': 'less',
|
| 80 |
+
'good': 'less',
|
| 81 |
+
'fine': 'less',
|
| 82 |
+
'worse': 'greater',
|
| 83 |
+
'worst': 'greater',
|
| 84 |
+
'lowest': 'greater',
|
| 85 |
+
'poor': 'greater',
|
| 86 |
+
'resolution': None,
|
| 87 |
+
'Γ₯': None,
|
| 88 |
+
'angstrom': None,
|
| 89 |
+
'than': None,
|
| 90 |
+
'under': 'less',
|
| 91 |
+
'below': 'less',
|
| 92 |
+
'above': 'greater',
|
| 93 |
+
'over': 'greater'
|
| 94 |
+
}
|
| 95 |
+
|
| 96 |
+
# Check if the original query mentions resolution
|
| 97 |
+
query_lower = query.lower()
|
| 98 |
+
|
| 99 |
+
# Determine resolution direction from query
|
| 100 |
+
for term, direction in resolution_terms.items():
|
| 101 |
+
if term in query_lower:
|
| 102 |
+
has_resolution_query = True
|
| 103 |
+
if direction: # if not None
|
| 104 |
+
resolution_direction = direction
|
| 105 |
+
|
| 106 |
+
# Also check for numerical values with Γ
|
| 107 |
+
if re.search(r'\d+\.?\d*\s*Γ₯?', query_lower):
|
| 108 |
+
has_resolution_query = True
|
| 109 |
+
|
| 110 |
+
# Clean and parse LLM response
|
| 111 |
+
for line in response.split('\n'):
|
| 112 |
+
if 'Resolution:' in line:
|
| 113 |
+
value = line.split('Resolution:')[1].strip()
|
| 114 |
+
if value.lower() not in ['none', 'n/a'] and has_resolution_query:
|
| 115 |
+
try:
|
| 116 |
+
# Extract just the number
|
| 117 |
+
res_value = ''.join(c for c in value if c.isdigit() or c == '.')
|
| 118 |
+
resolution_limit = float(res_value)
|
| 119 |
+
except ValueError:
|
| 120 |
+
pass
|
| 121 |
+
elif 'Method:' in line:
|
| 122 |
+
value = line.split('Method:')[1].strip()
|
| 123 |
+
if value.lower() not in ['none', 'n/a']:
|
| 124 |
+
method = value.upper()
|
| 125 |
+
elif 'Sequence:' in line:
|
| 126 |
+
value = line.split('Sequence:')[1].strip()
|
| 127 |
+
if value.lower() not in ['none', 'n/a']:
|
| 128 |
+
sequence = value
|
| 129 |
+
elif 'PDB_ID:' in line:
|
| 130 |
+
value = line.split('PDB_ID:')[1].strip()
|
| 131 |
+
if value.lower() not in ['none', 'n/a']:
|
| 132 |
+
pdb_id = value
|
| 133 |
+
|
| 134 |
+
# Build search query
|
| 135 |
+
queries = []
|
| 136 |
+
|
| 137 |
+
# Check if the query contains a protein sequence pattern
|
| 138 |
+
# Check for amino acid sequence (minimum 25 residues)
|
| 139 |
+
query_words = query.split()
|
| 140 |
+
for word in query_words:
|
| 141 |
+
# Check if the word consists of valid amino acid letters
|
| 142 |
+
if (len(word) >= 25 and # minimum 25 residues requirement
|
| 143 |
+
all(c in 'ACDEFGHIKLMNPQRSTVWY' for c in word.upper()) and
|
| 144 |
+
sum(c.isupper() for c in word) / len(word) > 0.8):
|
| 145 |
+
sequence = word
|
| 146 |
+
break
|
| 147 |
+
|
| 148 |
+
# If sequence is found, use SequenceQuery
|
| 149 |
+
if sequence:
|
| 150 |
+
if len(sequence) < 25:
|
| 151 |
+
print("Warning: Sequence must be at least 25 residues long. Skipping sequence search.")
|
| 152 |
+
sequence = None
|
| 153 |
+
else:
|
| 154 |
+
print(f"Adding sequence search with identity 100% for sequence: {sequence}")
|
| 155 |
+
sequence_query = SequenceQuery(
|
| 156 |
+
sequence,
|
| 157 |
+
identity_cutoff=1.0, # 100% identity
|
| 158 |
+
evalue_cutoff=1,
|
| 159 |
+
sequence_type="protein"
|
| 160 |
+
)
|
| 161 |
+
queries.append(sequence_query)
|
| 162 |
+
# If no sequence, proceed with text search
|
| 163 |
+
else:
|
| 164 |
+
# Clean the original query and add text search
|
| 165 |
+
clean_query = query.lower()
|
| 166 |
+
|
| 167 |
+
# Remove resolution numbers and terms if they exist
|
| 168 |
+
if has_resolution_query:
|
| 169 |
+
clean_query = re.sub(r'\d+\.?\d*\s*Γ₯?', '', clean_query)
|
| 170 |
+
for term in resolution_terms:
|
| 171 |
+
clean_query = clean_query.replace(term, '')
|
| 172 |
+
|
| 173 |
+
# Clean up extra spaces and trim
|
| 174 |
+
clean_query = ' '.join(clean_query.split())
|
| 175 |
+
|
| 176 |
+
print("Cleaned query:", clean_query)
|
| 177 |
+
|
| 178 |
+
# Add text search if query is not empty
|
| 179 |
+
if clean_query.strip():
|
| 180 |
+
text_query = AttributeQuery(
|
| 181 |
+
attribute="struct.title",
|
| 182 |
+
operator="contains_phrase",
|
| 183 |
+
value=clean_query
|
| 184 |
+
)
|
| 185 |
+
queries.append(text_query)
|
| 186 |
+
|
| 187 |
+
# Add resolution filter if specified
|
| 188 |
+
if resolution_limit and has_resolution_query:
|
| 189 |
+
operator = "less_or_equal" if resolution_direction == "less" else "greater_or_equal"
|
| 190 |
+
print(f"Adding resolution filter: {operator} {resolution_limit}Γ
")
|
| 191 |
+
resolution_query = AttributeQuery(
|
| 192 |
+
attribute="rcsb_entry_info.resolution_combined",
|
| 193 |
+
operator=operator,
|
| 194 |
+
value=resolution_limit
|
| 195 |
)
|
| 196 |
+
queries.append(resolution_query)
|
| 197 |
+
|
| 198 |
+
# Add PDB ID search if specified
|
| 199 |
+
if pdb_id:
|
| 200 |
+
print(f"Searching for specific PDB ID: {pdb_id}")
|
| 201 |
+
id_query = AttributeQuery(
|
| 202 |
+
attribute="rcsb_id",
|
| 203 |
+
operator="exact_match",
|
| 204 |
+
value=pdb_id.upper()
|
| 205 |
+
)
|
| 206 |
+
queries = [id_query] # Override other queries for direct PDB ID search
|
| 207 |
+
|
| 208 |
+
# Add experimental method filter if specified
|
| 209 |
+
if method:
|
| 210 |
+
print(f"Adding experimental method filter: {method}")
|
| 211 |
+
method_query = AttributeQuery(
|
| 212 |
+
attribute="exptl.method",
|
| 213 |
+
operator="exact_match",
|
| 214 |
+
value=method
|
| 215 |
+
)
|
| 216 |
+
queries.append(method_query)
|
| 217 |
+
|
| 218 |
+
# Combine queries with AND operator
|
| 219 |
+
if queries:
|
| 220 |
+
final_query = queries[0]
|
| 221 |
+
for q in queries[1:]:
|
| 222 |
+
final_query = final_query & q
|
| 223 |
+
|
| 224 |
+
print("Final query:", final_query)
|
| 225 |
+
|
| 226 |
+
# Execute search
|
| 227 |
+
session = final_query.exec()
|
| 228 |
+
results = []
|
| 229 |
+
|
| 230 |
+
# Process results safely with additional information
|
| 231 |
+
try:
|
| 232 |
+
for entry in session:
|
| 233 |
+
# Handle both string and object types
|
| 234 |
+
if isinstance(entry, str):
|
| 235 |
+
result = {
|
| 236 |
+
'PDB ID': entry
|
| 237 |
+
}
|
| 238 |
+
else:
|
| 239 |
+
# Handle object type
|
| 240 |
+
result = {
|
| 241 |
+
'PDB ID': entry.identifier
|
| 242 |
+
}
|
| 243 |
+
|
| 244 |
+
results.append(result)
|
| 245 |
+
except Exception as e:
|
| 246 |
+
print(f"Error processing results: {str(e)}")
|
| 247 |
+
# If error occurs during processing, at least return PDB IDs
|
| 248 |
+
if isinstance(entry, str):
|
| 249 |
+
results.append({'PDB ID': entry})
|
| 250 |
+
|
| 251 |
+
print(f"Found {len(results)} structures")
|
| 252 |
+
return results
|
| 253 |
+
|
| 254 |
+
return []
|
| 255 |
+
|
| 256 |
+
except Exception as e:
|
| 257 |
+
print(f"Error during search: {str(e)}")
|
| 258 |
+
print(f"Error type: {type(e)}")
|
| 259 |
+
return []
|
| 260 |
|
| 261 |
+
def pdbsummary(name):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 262 |
|
| 263 |
+
search_engine = ProteinSearchEngine()
|
| 264 |
|
| 265 |
+
query = ProteinQuery(
|
| 266 |
+
name,
|
| 267 |
+
max_resolution= 5.0
|
| 268 |
+
)
|
|
|
|
|
|
|
|
|
|
| 269 |
|
| 270 |
+
results = search_engine.search(query)
|
|
|
|
|
|
|
|
|
|
|
|
|
| 271 |
|
| 272 |
+
answer = ""
|
| 273 |
+
for i, structure in enumerate(results, 1):
|
| 274 |
+
answer += f"\n{i}. PDB ID : {structure.pdb_id}\n"
|
| 275 |
+
answer += f"\nResolution : {structure.resolution:.2f} A \n"
|
| 276 |
+
answer += f"Method : {structure.method}\n Title : {structure.title}\n"
|
| 277 |
+
answer += f"Release Date : {structure.release_date}\n Sequence length: {len(structure.sequence)} aa\n"
|
| 278 |
+
answer += f" Sequence:\n {structure.sequence}\n"
|
| 279 |
|
| 280 |
+
return answer
|
| 281 |
|
| 282 |
+
def create_interactive_table(df):
|
| 283 |
+
if df.empty:
|
| 284 |
+
return go.Figure()
|
| 285 |
+
|
| 286 |
+
# Create interactive table
|
| 287 |
+
table = go.Figure(data=[go.Table(
|
| 288 |
+
header=dict(
|
| 289 |
+
values=list(df.columns),
|
| 290 |
+
fill_color='paleturquoise',
|
| 291 |
+
align='left',
|
| 292 |
+
font=dict(size=14),
|
| 293 |
+
),
|
| 294 |
+
cells=dict(
|
| 295 |
+
values=[df[col] for col in df.columns],
|
| 296 |
+
align='left',
|
| 297 |
+
font=dict(size=13),
|
| 298 |
+
height=30
|
| 299 |
+
),
|
| 300 |
+
columnwidth=[len(str(max(df[col], key=len))) for col in df.columns]
|
| 301 |
+
)])
|
| 302 |
+
|
| 303 |
+
# Update table layout
|
| 304 |
+
table.update_layout(
|
| 305 |
+
margin=dict(l=0, r=0, t=0, b=0),
|
| 306 |
+
height=400,
|
| 307 |
+
autosize=True
|
| 308 |
+
)
|
| 309 |
+
|
| 310 |
+
return table
|
| 311 |
|
| 312 |
+
# Simplified Shiny app UI definition
|
| 313 |
+
app_ui = ui.page_fluid(
|
| 314 |
+
ui.tags.head(
|
| 315 |
+
ui.tags.style("""
|
| 316 |
+
.table a {
|
| 317 |
+
color: #0d6efd;
|
| 318 |
+
text-decoration: none;
|
| 319 |
+
}
|
| 320 |
+
.table a:hover {
|
| 321 |
+
color: #0a58ca;
|
| 322 |
+
text-decoration: underline;
|
| 323 |
+
}
|
| 324 |
+
""")
|
| 325 |
+
),
|
| 326 |
+
ui.h2("Advanced PDB Structure Search Tool"),
|
| 327 |
+
ui.row(
|
| 328 |
+
ui.column(12,
|
| 329 |
+
ui.input_text("query", "Search Query",
|
| 330 |
+
value="Human insulin"),
|
| 331 |
+
)
|
| 332 |
+
),
|
| 333 |
+
ui.row(
|
| 334 |
+
ui.column(12,
|
| 335 |
+
ui.p("Example queries:"),
|
| 336 |
+
ui.tags.ul(
|
| 337 |
+
ui.tags.li("Human hemoglobin C resolution better than 2.5Γ
"),
|
| 338 |
+
ui.tags.li("Find structures containing sequence MNIFEMLRIDEGLRLKIYKDTEGYYTIGIGHLLTKSPSLNAAKSELDKAIGRNTNGVITKDEAEKLFNQDVDAAVRGILRNAKLKPVYDSLDAVRRAALINMVFQMGETGVAGFTNSLRMLQQKRWDEAAVNLAKSRWYNQTPNRAKRVITTFRTGTWDAYKNL"),
|
| 339 |
|
| 340 |
+
),
|
| 341 |
+
)
|
| 342 |
+
),
|
| 343 |
+
ui.row(
|
| 344 |
+
ui.column(12,
|
| 345 |
+
ui.input_action_button("search", "Search", class_="btn-primary"),
|
| 346 |
+
)
|
| 347 |
+
),
|
| 348 |
+
ui.row(
|
| 349 |
+
ui.column(12,
|
| 350 |
+
ui.h4("Search Parameters:"),
|
| 351 |
+
ui.output_text("search_conditions"),
|
| 352 |
+
)
|
| 353 |
+
),
|
| 354 |
+
ui.row(
|
| 355 |
+
ui.column(12,
|
| 356 |
+
ui.h4("Top 10 Results:"),
|
| 357 |
+
output_widget("results_table"),
|
| 358 |
+
ui.download_button("download", "Download Results")
|
| 359 |
+
)
|
| 360 |
+
)
|
| 361 |
+
)
|
| 362 |
|
| 363 |
+
def server(input, output, session):
|
| 364 |
+
assistant = PDBSearchAssistant()
|
| 365 |
+
results_store = reactive.Value([])
|
| 366 |
+
|
| 367 |
+
@reactive.Effect
|
| 368 |
+
@reactive.event(input.search)
|
| 369 |
+
def _():
|
| 370 |
+
results = assistant.search_pdb(query=input.query())
|
| 371 |
+
results_store.set(results)
|
| 372 |
+
|
| 373 |
+
# Convert results to DataFrame and add hyperlinks
|
| 374 |
+
df = pd.DataFrame(results)
|
| 375 |
+
if not df.empty:
|
| 376 |
+
df['PDB ID'] = df['PDB ID'].apply(
|
| 377 |
+
lambda x: f'<a href="https://www.rcsb.org/3d-view/{x}" target="_blank">{x}</a>'
|
| 378 |
+
)
|
| 379 |
+
|
| 380 |
+
@output
|
| 381 |
+
@render_widget
|
| 382 |
+
def results_table():
|
| 383 |
+
return create_interactive_table(df) # id μμΌλ‘ μ λ ¬λλκ±°μΈλ― Top rank μμ μλ
|
| 384 |
+
|
| 385 |
+
@output
|
| 386 |
+
@render.text
|
| 387 |
+
def search_conditions():
|
| 388 |
+
results = results_store.get()
|
| 389 |
+
return f"""
|
| 390 |
+
Applied Search Conditions:
|
| 391 |
+
- Query: {input.query()}
|
| 392 |
+
- Total structures found: {len(results)}
|
| 393 |
+
"""
|
| 394 |
+
|
| 395 |
+
@output
|
| 396 |
+
@render.download(filename="pdb_search_results.csv")
|
| 397 |
+
def download():
|
| 398 |
+
df = pd.DataFrame(results_store.get())
|
| 399 |
+
return df.to_csv(index=False)
|
| 400 |
|
| 401 |
+
app = App(app_ui, server)
|
| 402 |
|
| 403 |
+
if __name__ == "__main__":
|
| 404 |
+
import nest_asyncio
|
| 405 |
+
nest_asyncio.apply()
|
| 406 |
+
app.run(port=8000)
|
|
|
requirements.txt
CHANGED
|
@@ -1,6 +1,9 @@
|
|
| 1 |
-
|
|
|
|
|
|
|
| 2 |
shiny
|
| 3 |
-
shinywidgets
|
| 4 |
-
plotly
|
| 5 |
pandas
|
| 6 |
-
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
transformers
|
| 2 |
+
rcsbsearchapi
|
| 3 |
+
python-dotenv
|
| 4 |
shiny
|
|
|
|
|
|
|
| 5 |
pandas
|
| 6 |
+
plotly
|
| 7 |
+
shinywidgets
|
| 8 |
+
nest-asyncio
|
| 9 |
+
torch
|