Skip to main content
Back to challenges
Challenge

Nanopore Long-Read Structural Variant Indexer with LlamaIndex and AI21 Studio

Oxford Nanopore long-read sequencing unlocks structural variants across biopharma workflows. You will build a RAG pipeline using LlamaIndex and AI21 Studio to query structural variant databases and genomic documentation, achieving an explicit query retrieval precision of over 88%.

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.

Build a genomic RAG system using LlamaIndex and AI21 Studio to index and query Nanopore long-read structural variants.

Delivery guide

How work is evaluated

Evaluation

Evaluates genomic RAG retrieval precision and answer factual precision on Nanopore SV queries.

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

variant_id_match

Ensure correct structural variant ID is returned

Binary check

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

Dimension 2

retrieval_precision

Precision of relevant genomic variants retrieved • 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

What you should walk away with

  • Construct knowledge indices over VCF genomic files using LlamaIndex framework

  • Integrate AI21 Studio language models for specialized biological information retrieval

  • Execute natural language queries for structural variant functional impact identification

Resources and assets

Reference links and supporting material

Dataset notes

Oxford Nanopore genomic structural variant VCF dataset with 1,000 annotated biopharma clinical samples.

Challenge setup

Evaluates genomic RAG retrieval precision and answer factual precision on Nanopore SV queries.

Challenge input

JSON object with query_string

LlamaIndex

Mandatory framework for biological data retrieval.

AI21 Studio

Provides language model reasoning capabilities.

Evaluated output

JSON object with answer string and retrieved_variant_ids list

Evaluation checks
  • Ensure correct structural variant ID is returned
  • Precision of relevant genomic variants retrieved • target: 0.88 • range: 0-1
Configured evaluation evidence
  • Benchmark: GenoRAG-Bench
  • Retrieval Precision target: 0.88
  • 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
  • AI21 Studio · Required
  • Llama Index · Optional
  • Exa · Optional

Evaluation contract

  • variant_id_match · Weight 1
  • retrieval_precision · Weight 1

Metrics

  • retrieval_precision · Target 0.88

Benchmarks

  • GenoRAG-Bench
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with query_string

    Output: JSON object with answer string and retrieved_variant_ids list

    Execution harness: Python

Evaluation tools

  • LlamaIndex · RAG Framework
  • AI21 Studio · Language Model Endpoint

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Nanopore Long-Read Structural Variant Indexer with LlamaIndex and AI21 Studio