Protein Functional Annotation Pipeline
Build a high-performance protein annotation agent using the Google ADK (Agent Development Kit). The agent's core responsibility is to reconcile unreviewed protein sequences with reviewed UniProtKB entries. You will deploy a sequence alignment service on Modal to handle heavy BLAST-like computations or HMMER searches. The Google ADK agent will orchestrate the flow: receiving an unreviewed sequence, triggering a Modal inference job for alignment, and then fetching metadata from UniProt (taxonomy, active sites, PTMs) to explain the functional confidence. The agent must handle the multimodal aspect of protein data (sequences and structural metadata) and return a detailed 'Evidence Reconciliation' report that identifies conserved domains and functional residues.
What you are building
The core problem, expected build, and operating context for this challenge.
Orchestrate protein sequence alignment on Modal and functional metadata retrieval from UniProt using a Google ADK-powered agent.
How work is evaluated
Evaluation will check the agent's ability to correctly transfer annotations from a reviewed template to an unreviewed query sequence.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
Modal Execution
Verifies that the alignment was actually performed on Modal serverless.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Annotation Accuracy
Percentage of correctly identified functional residues. • target: 0.95 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Develop a multimodal protein analysis agent using Google ADK and Gemini
Deploy and scale bioinformatics sequence alignment workloads on Modal's serverless infrastructure
Reconcile reviewed (Swiss-Prot) and unreviewed (TrEMBL) UniProt records programmatically
Implement a structured annotation pipeline that maps PTMs and active sites from templates to query sequences
Reference links and supporting material
Sample data for 1 tasks
Challenge setup
Evaluation will check the agent's ability to correctly transfer annotations from a reviewed template to an unreviewed query sequence.
Challenge input
Unreviewed FASTA sequence
Modal
Serverless cloud for AI workloads.
Google AI model provider
Google ADK
Agent Development Kit for
Evaluated output
JSON mapping of residue positions to functions
- Verifies that the alignment was actually performed on Modal serverless.
- Percentage of correctly identified functional residues. • target: 0.95 • range: 0-1
- Annotation Accuracy target: 0.95
- 1 public reference case
- Python execution harness
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
- Modal · Optional
- Google · Required
- Google ADK · Optional
- Modal · Optional
- Modal · Optional
Evaluation contract
- Modal Execution · Weight 1
- Annotation Accuracy · Weight 1
Metrics
- Annotation Accuracy · Target 0.95
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: Unreviewed FASTA sequence
Output: JSON mapping of residue positions to functions
Execution harness: Python
Recipe freeze marker
No recipe freeze time is recorded.