Skip to main content
Loading navigation
Back to challenges

CrewAI Tumor Mutant Dosage Modeling on Modal Serverless

Analyzing mutant dosage patterns in massive clinical cancer cohorts requires distributed compute and coordinated agent roles. Build a CrewAI agent team deployed on Modal serverless infrastructure to process genomic allele fractions and correlate dosage with metastatic tropism. Achieve end-to-end cohort processing within 15 minutes.

Data ScienceHosted by Vera
Sign in free to start

What you are building

Orchestrate multi-agent cancer genomics analysis using CrewAI role-playing agents deployed on Modal serverless GPUs for parallel mutation dosage calculation.

How work is evaluated

Evaluation

Tests execution speed, dosage calculation correctness, and agent output report quality on synthetic cancer cohort matrices.

Datasets

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

How submissions are scored

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

modal_execution_success

Verifies that Modal function executes without runtime exceptions.

Binary check

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

Dimension 2

correlation_precision

Precision of identified gene-site tropism correlations. • target: 0.88 • 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

  • Define specialized CrewAI roles: Genomics Analyst, Biostatistician, and Clinical Reporter

  • Deploy scalable parallel inference workflows using Modal serverless functions

  • Calculate mutant allele dosage and variant allele fraction (VAF) from cohort matrices

  • Synthesize organ-specific metastatic tropism association reports

Resources

Dataset notes

Synthetic clinical cancer cohort dataset derived from genomic patterns representing allele fractions, copy numbers, and metastatic site labels.

Challenge setup

Tests execution speed, dosage calculation correctness, and agent output report quality on synthetic cancer cohort matrices.

Challenge input

CSV dataset path with sample_id, gene, VAF, CNV, and metastatic_site

CrewAI

Required framework for coordinating specialized agent teams.

Modal

Provides fast cloud infrastructure for genomic data processing.

Evaluated output

JSON containing top gene-site associations and hazard ratios

Evaluation checks
  • Verifies that Modal function executes without runtime exceptions.
  • Precision of identified gene-site tropism correlations. • target: 0.88 • range: 0-1
Configured evaluation evidence
  • Benchmark: OncoDosage-Bench
  • Correlation Precision target: 0.88
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Environment
  • Modal · Required
Action Space
  • CrewAI · Optional
  • Lingo · Optional
Policy Serving
  • Modal · Required

Evaluation contract

  • modal_execution_success · Weight 1
  • correlation_precision · Weight 1

Metrics

  • correlation_precision · Target 0.88

Benchmarks

  • OncoDosage-Bench
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: CSV dataset path with sample_id, gene, VAF, CNV, and metastatic_site

    Output: JSON containing top gene-site associations and hazard ratios

    Execution harness: Python

Evaluation tools

  • CrewAI · Role-playing agent orchestrator
  • Modal · Serverless compute runtime

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about CrewAI Tumor Mutant Dosage Modeling on Modal Serverless