Skip to main content
Back to challenges
Challenge

Multi-Agent MicroRNA Biomarker Discovery for Cancer Trials with AutoGen and Claude 4 Sonnet

Early detection of pancreatic cancer relies on non-invasive microRNA profiling from urine samples. Build a multi-agent AutoGen system powered by Claude 4 Sonnet to analyze microRNA expression matrices and classify high-risk pancreatic cancer cohorts with an AUC greater than 0.90.

Data ScienceHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Develop a multi-agent collaborative biomarker discovery engine using AutoGen and Claude 4 Sonnet.

Delivery guide

How work is evaluated

Evaluation

Evaluates pancreatic cancer cohort classification ROC-AUC performance.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

output_range_check

Checks if cancer probability is bounded between 0.0 and 1.0.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

roc_auc

Area under the ROC curve on holdout test samples. • target: 0.9 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Implement multi-agent conversation workflows using Microsoft AutoGen.

  • Utilize Claude 4 Sonnet for advanced biological feature selection reasoning.

  • Train classification models on high-dimensional urine microRNA profiling data.

  • Validate biomarker diagnostic thresholds against clinical ground truth.

Resources and assets

Reference links and supporting material

Dataset notes

Normalized microRNA expression profiles from simulated urine samples of healthy and pancreatic cancer patients.

Challenge setup

Evaluates pancreatic cancer cohort classification ROC-AUC performance.

Challenge input

JSON containing patient miRNA expression levels.

AutoGen

Required multi-agent orchestrator for interactive feature selection.

Claude 4 Sonnet

Provides bio-reasoning capabilities for microRNA pathway analysis.

Evaluated output

JSON containing predicted risk probability and risk tier.

Evaluation checks
  • Checks if cancer probability is bounded between 0.0 and 1.0.
  • Area under the ROC curve on holdout test samples. • target: 0.9 • range: 0-1
Configured evaluation evidence
  • Benchmark: Biomarker MicroRNA Classification Challenge
  • Roc Auc target: 0.9
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

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

Action Space
  • AutoGen · Optional
  • Aembit · Optional
Policy Serving
  • Claude 4 Sonnet · Required

Evaluation contract

  • output_range_check · Weight 1
  • roc_auc · Weight 1

Metrics

  • roc_auc · Target 0.9

Benchmarks

  • Biomarker MicroRNA Classification Challenge
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing patient miRNA expression levels.

    Output: JSON containing predicted risk probability and risk tier.

    Execution harness: Python

Evaluation tools

  • AutoGen · Multi-agent Framework
  • Claude 4 Sonnet · LLM Engine

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Multi-Agent MicroRNA Biomarker Discovery for Cancer Trials with AutoGen and Claude 4 Sonnet