Setup¶
import os
from pathlib import Path
from IPython.display import Image, display, HTML
from noctua.barista import BaristaClient
from noctua.noctua_screenshots import NoctuaScreenshotCapture
# Ensure BARISTA_TOKEN is set in your environment
# For development: Contact the GO team for a dev token
# For production: Get token from Noctua login
if not os.environ.get("BARISTA_TOKEN"):
print("ERROR: BARISTA_TOKEN environment variable not set")
print("Please set: export BARISTA_TOKEN=your-token-here")
raise SystemExit(1)
# Create screenshots directory
screenshots_dir = Path("screenshots")
screenshots_dir.mkdir(exist_ok=True)
print("Environment configured")
print(f"Screenshots will be saved to: {screenshots_dir.absolute()}")
Create Model and Initialize Screenshot Capture¶
We'll create a new model and open it in Noctua for visual tracking.
# Create Barista client
client = BaristaClient()
# Create a new model
response = client.create_model(title="Visual Demo: Signal Transduction Pathway")
if not response.ok:
print(f"Failed to create model: {response.raw}")
raise Exception("Could not create model")
model_id = response.model_id
print(f"Created model: {model_id}")
# Get the Noctua URL with token
token = os.environ.get("BARISTA_TOKEN", "")
noctua_url = f"http://noctua-dev.berkeleybop.org/editor/graph/{model_id}"
if token:
noctua_url += f"?barista_token={token}"
print(f"\nNoctua URL: {noctua_url}")
display(HTML(f'<a href="{noctua_url}" target="_blank">Open in Noctua</a>'))
Initialize Screenshot Capture¶
This will open a browser window to track changes. Set headless=True to run without GUI.
# Initialize screenshot capture
# Note: Set headless=True to run without opening a browser window
capture = NoctuaScreenshotCapture(
headless=False, # Set to True for headless mode
screenshot_dir="screenshots",
dev_mode=True
)
try:
capture.start()
print("Browser started successfully")
# Open the model in Noctua
if capture.open_model(model_id, wait_time=8):
print("Model opened in browser")
# Take initial screenshot
initial_screenshot = capture.capture(
filename="01_empty_model.png",
description="Empty model"
)
if initial_screenshot:
print("Initial screenshot saved")
display(Image(initial_screenshot, width=800))
else:
print("Could not open model in browser")
print("Make sure Chrome/Firefox and appropriate driver are installed")
except Exception as e:
print(f"Error setting up browser: {e}")
print("Continuing without screenshots...")
capture.close()
capture = None
Step 1: Add Molecular Functions¶
Add molecular function nodes and capture screenshots after each addition.
# Store individual IDs for later use
individuals = {}
# Add receptor activity
print("Adding receptor activity...")
response = client.add_individual(
model_id,
"GO:0004888", # transmembrane signaling receptor activity
assign_var="receptor"
)
if response.ok:
# Find the added individual
for ind in response.individuals:
types = ind.get("type", [])
for t in types:
if t.get("id") == "GO:0004888":
individuals["receptor"] = ind["id"]
print(f"Added receptor: {ind['id']}")
break
# Refresh and screenshot
if capture:
capture.refresh_model(wait_time=3)
screenshot = capture.capture(
filename="02_receptor_added.png",
description="After adding receptor"
)
if screenshot:
display(Image(screenshot, width=800))
# Add GTPase activity
print("Adding GTPase activity...")
response = client.add_individual(
model_id,
"GO:0003924", # GTPase activity
assign_var="gtpase"
)
if response.ok:
for ind in response.individuals:
types = ind.get("type", [])
for t in types:
if t.get("id") == "GO:0003924":
individuals["gtpase"] = ind["id"]
print(f"Added GTPase: {ind['id']}")
break
if capture:
capture.refresh_model(wait_time=3)
screenshot = capture.capture(
filename="03_gtpase_added.png",
description="After adding GTPase"
)
if screenshot:
display(Image(screenshot, width=800))
# Add kinase activity
print("Adding kinase activity...")
response = client.add_individual(
model_id,
"GO:0004674", # protein serine/threonine kinase activity
assign_var="kinase"
)
if response.ok:
for ind in response.individuals:
types = ind.get("type", [])
for t in types:
if t.get("id") == "GO:0004674":
individuals["kinase"] = ind["id"]
print(f"Added kinase: {ind['id']}")
break
if capture:
capture.refresh_model(wait_time=3)
screenshot = capture.capture(
filename="04_kinase_added.png",
description="After adding kinase"
)
if screenshot:
display(Image(screenshot, width=800))
Step 2: Add Causal Relationships¶
Connect the molecular functions with regulatory relationships.
# Add edge: Receptor -> GTPase
if "receptor" in individuals and "gtpase" in individuals:
print("Adding edge: Receptor -> GTPase")
response = client.add_fact(
model_id,
subject_id=individuals["receptor"],
object_id=individuals["gtpase"],
predicate_id="RO:0002413" # directly positively regulates
)
if response.ok:
print("Edge added successfully")
if capture:
capture.refresh_model(wait_time=3)
screenshot = capture.capture(
filename="05_receptor_to_gtpase.png",
description="After adding Receptor->GTPase edge"
)
if screenshot:
display(Image(screenshot, width=800))
# Add edge: GTPase -> Kinase
if "gtpase" in individuals and "kinase" in individuals:
print("Adding edge: GTPase -> Kinase")
response = client.add_fact(
model_id,
subject_id=individuals["gtpase"],
object_id=individuals["kinase"],
predicate_id="RO:0002413" # directly positively regulates
)
if response.ok:
print("Edge added successfully")
if capture:
capture.refresh_model(wait_time=3)
screenshot = capture.capture(
filename="06_gtpase_to_kinase.png",
description="After adding GTPase->Kinase edge"
)
if screenshot:
display(Image(screenshot, width=800))
Step 3: Add Evidence¶
Add evidence annotation to support the relationships.
# Add evidence to GTPase->Kinase edge
if "gtpase" in individuals and "kinase" in individuals:
print("Adding evidence to GTPase->Kinase edge...")
evidence_requests = client.req_add_evidence_to_fact(
model_id,
subject_id=individuals["gtpase"],
object_id=individuals["kinase"],
predicate_id="RO:0002413",
eco_id="ECO:0000314", # direct assay evidence
sources=["PMID:12345678", "PMID:87654321"],
with_from=["UniProtKB:P01112"] # HRAS
)
response = client.m3_batch(evidence_requests)
if response.ok:
print("Evidence added successfully")
if capture:
capture.refresh_model(wait_time=3)
screenshot = capture.capture(
filename="07_evidence_added.png",
description="After adding evidence"
)
if screenshot:
display(Image(screenshot, width=800))
Step 4: View Complete Model¶
Final screenshot of the complete model.
# Get final model state
response = client.get_model(model_id)
if response.ok:
print("Final model state:")
print(f"- Model ID: {response.model_id}")
print(f"- Individuals: {len(response.individuals)}")
print(f"- Facts: {len(response.facts)}")
if capture:
# Take a final screenshot
screenshot = capture.capture(
filename="08_complete_model.png",
description="Complete model"
)
if screenshot:
print("\nFinal model visualization:")
display(Image(screenshot, width=800))
Generate Screenshot Gallery¶
Create an HTML gallery of all screenshots taken during the session.
# Generate gallery of all screenshots
from pathlib import Path
screenshot_files = sorted(Path("screenshots").glob("*.png"))
if screenshot_files:
html_content = "<h3>Screenshot Gallery</h3>\n"
html_content += "<div style="display: flex; flex-wrap: wrap;">\n"
for img_path in screenshot_files:
html_content += f'''
<div style="margin: 10px; text-align: center;">
<img alt="No description has been provided for this image" src="{img_path}" style="border: 1px solid #ddd;" width="300">
<p style="font-size: 12px;">{img_path.name}</p>
</img></div>
'''
html_content += "</div>\n"
display(HTML(html_content))
else:
print("No screenshots found")
Screenshot Gallery¶
01_empty_model.png
02_receptor_added.png
03_gtpase_added.png
04_kinase_added.png
05_receptor_to_gtpase.png
06_gtpase_to_kinase.png
07_evidence_added.png
08_complete_model.png
Cleanup¶
Summary¶
This notebook demonstrated:
- Automated Browser Control: Using Selenium to open and control Noctua
- Visual Documentation: Capturing screenshots at each step of model construction
- API + UI Sync: Showing how API changes appear in the Noctua interface
- Complete Workflow: From empty model to fully annotated pathway with evidence
The screenshots provide visual documentation of the model building process, useful for: - Training materials - Documentation - Debugging model construction - Quality assurance
Tips¶
- Use
headless=Truefor automated pipelines - Adjust wait times based on network speed
- Screenshots are automatically saved with descriptive names
- The browser window can be controlled programmatically for zooming, panning, etc.