Skip to main content
Back to challenges
Challenge

Somatic Mutation Dosage RAG and Metastatic Tropism Agent using LlamaIndex

Clinical cancer interpretation requires mapping gene mutant dosages across tens of thousands of tumor samples to predict metastatic organ tropism. Build an advanced RAG agent using LlamaIndex to query somatic mutation databases and output high-confidence metastatic tropism predictions with 92% evaluation score accuracy.

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 an advanced genomic RAG system using LlamaIndex to map mutant dosage to metastatic organ tropism.

Delivery guide

How work is evaluated

Evaluation

Evaluates RAG retrieval accuracy for predicting metastatic organ tropism based on somatic mutation dosage.

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

tropism_organ_validity

Verifies organ site belongs to standard clinical anatomical list.

Binary check

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

Dimension 2

rag_eval_score

Evaluation score based on prediction correctness and retrieval context relevance. • target: 0.92 • 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 advanced retrieval-augmented generation (RAG) using LlamaIndex.

  • Index large-scale clinical cancer genomic datasets (60,000 sample cohort profiles).

  • Query somatic mutation dosage vectors to infer metastatic organ tropism probabilities.

  • Achieve 92% evaluation accuracy on metastatic target site prediction.

Resources and assets

Reference links and supporting material

Dataset notes

Synthetic somatic mutation dosage profiles across 60,000 tumor samples with associated organ tropism annotations.

Challenge setup

Evaluates RAG retrieval accuracy for predicting metastatic organ tropism based on somatic mutation dosage.

Challenge input

JSON containing somatic mutations and copy number/dosage ratios.

LlamaIndex

Leading framework for indexing and querying structured/unstructured enterprise data.

Evaluated output

JSON containing target_organ_tropism and confidence_score.

Evaluation checks
  • Verifies organ site belongs to standard clinical anatomical list.
  • Evaluation score based on prediction correctness and retrieval context relevance. • target: 0.92 • range: 0-1
Configured evaluation evidence
  • Benchmark: Oncology RAG & Tropism Prediction Benchmark
  • Rag Eval Score target: 0.92
  • 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
  • Meta · Required
  • Llama Index · Optional
  • Testers AI · Optional

Evaluation contract

  • tropism_organ_validity · Weight 1
  • rag_eval_score · Weight 1

Metrics

  • rag_eval_score · Target 0.92

Benchmarks

  • Oncology RAG & Tropism Prediction Benchmark
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing somatic mutations and copy number/dosage ratios.

    Output: JSON containing target_organ_tropism and confidence_score.

    Execution harness: Python

Evaluation tools

  • LlamaIndex · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Somatic Mutation Dosage RAG and Metastatic Tropism Agent using LlamaIndex