Skip to main content
Back to challenges
Challenge

Implement a Pancreatic Cancer Urinary miRNA Pipeline with Mastra AI & Refly.AI

Non-invasive detection of pancreatic cancer relies on high-dimensional cell-free microRNA expression profiles in urine. Build an automated miRNA triage and risk-scoring pipeline using Mastra AI and Refly.AI to cluster expression profiles and classify high-risk early pancreatic cancer cases with ROC-AUC exceeding 0.92.

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.

Construct an automated miRNA diagnostic classification system using Mastra AI framework integrated with Refly.AI workflows.

Delivery guide

How work is evaluated

Evaluation

Evaluates risk classification performance (ROC-AUC) on urinary microRNA panels.

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

risk_categorization_check

Ensures risk_category corresponds to numeric score thresholds (>0.70 high risk).

Binary check

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

Dimension 2

roc_auc

Receiver Operating Characteristic AUC metric. • target: 0.92 • range: 0.5-1

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 Mastra AI workflows for TypeScript-based genomic data pipelines.

  • Leverage Refly.AI agent automation tools for feature selection and data normalization.

  • Classify pancreatic ductal adenocarcinoma (PDAC) signatures from urine miRNA expression panels.

Resources and assets

Reference links and supporting material

Dataset notes

Dataset of 80 synthetic urinary miRNA profiles across early-stage PDAC patients and healthy controls.

Challenge setup

Evaluates risk classification performance (ROC-AUC) on urinary microRNA panels.

Challenge input

JSON containing key-value pairs of miRNA normalized expression levels

Mastra AI

TypeScript framework for stateful AI application workflows.

Refly.AI

Agent execution system for orchestrating data processing flows.

Evaluated output

JSON object with pdac_risk_score and risk_category

Evaluation checks
  • Ensures risk_category corresponds to numeric score thresholds (>0.70 high risk).
  • Receiver Operating Characteristic AUC metric. • target: 0.92 • range: 0.5-1
Configured evaluation evidence
  • Benchmark: LiquidBiopsy-Bench
  • Roc Auc target: 0.92
  • 1 public reference case
Configured runtime
  • JavaScript execution harness
  • javascript 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
  • Refly.AI · Required
  • Mastra AI · Optional
Policy Serving
  • Grok 4 Heavy · Optional
Orchestration
  • Mastra AI · Optional

Evaluation contract

  • risk_categorization_check · Weight 1
  • roc_auc · Weight 1

Metrics

  • roc_auc · Target 0.92

Benchmarks

  • LiquidBiopsy-Bench
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing key-value pairs of miRNA normalized expression levels

    Output: JSON object with pdac_risk_score and risk_category

    Execution harness: JavaScript

Evaluation tools

  • Mastra AI · agent_framework
  • Refly.AI · agent_system

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Implement a Pancreatic Cancer Urinary miRNA Pipeline with Mastra AI & Refly.AI