Skip to main content
Loading navigation
Back to challenges

Cell-Free DNA Drug Safety Monitoring Agent with OpenAI Agents SDK

Monitoring drug-induced tissue toxicity requires tracking cell-free DNA tissue-of-origin signals. You will build an agent using OpenAI Agents SDK with function calling to evaluate cfDNA methylation signatures in liquid biopsy logs, flagging organ toxicity warnings with 95% sensitivity.

Data ScienceHosted by Vera
Sign in free to start

What you are building

Build an OpenAI Agents SDK system with tool calling to monitor organ-specific cell-free DNA methylation signals for drug safety.

How work is evaluated

Evaluation

Evaluates toxicity detection accuracy and sensitivity on cell-free DNA methylation time-series data.

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

critical_alert_check

Ensure critical liver toxicity fold change triggers appropriate alert level

Binary check

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

Dimension 2

sensitivity

Sensitivity in detecting true organ injury events • target: 0.95 • 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

  • Implement autonomous agent loops using official OpenAI Agents SDK

  • Define type-safe python function tools for cfDNA deconvolution analysis

  • Detect organ-specific acute cell death signatures during Phase 1 drug trials

Resources

Dataset notes

Cell-free DNA methylation fraction dataset representing 150 clinical trial subjects across multiple drug dosing cohorts.

Challenge setup

Evaluates toxicity detection accuracy and sensitivity on cell-free DNA methylation time-series data.

Challenge input

JSON object with liver_fraction and kidney_fraction floats

OpenAI Agents SDK

Mandatory OpenAI agent SDK framework.

Evaluated output

JSON object with alert_level string and affected_organs list

Evaluation checks
  • Ensure critical liver toxicity fold change triggers appropriate alert level
  • Sensitivity in detecting true organ injury events • target: 0.95 • range: 0-1
Configured evaluation evidence
  • Benchmark: ToxTriage-Bench
  • Sensitivity target: 0.95
  • 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

Action Space
  • OpenAI · Required
  • Durable · Optional
  • Text Generation Inference · Optional

Evaluation contract

  • critical_alert_check · Weight 1
  • sensitivity · Weight 1

Metrics

  • sensitivity · Target 0.95

Benchmarks

  • ToxTriage-Bench
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with liver_fraction and kidney_fraction floats

    Output: JSON object with alert_level string and affected_organs list

    Execution harness: Python

Evaluation tools

  • OpenAI Agents SDK · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Cell-Free DNA Drug Safety Monitoring Agent with OpenAI Agents SDK