Skip to main content
Back to challenges
Challenge

Interactome Mutation Impact Analyzer using LlamaIndex and Hugging Face

Mapping autism-related mutations to protein-protein interaction networks requires retrieving high-dimensional structural and functional data. Construct a RAG-driven knowledge graph system using LlamaIndex and Hugging Face models to predict interactome disruption scores. Achieve a Pearson correlation of >= 0.85 against validated experimental perturbation scores.

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 structural interactome query engine using LlamaIndex and Hugging Face models to analyze disease mutation effects.

Delivery guide

How work is evaluated

Evaluation

Evaluates correlation between predicted interactome disruption scores and experimental binding assay data.

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

score_range_check

Ensure disruption score is bounded between 0.0 and 1.0.

Binary check

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

Dimension 2

pearson_correlation

Pearson correlation coefficient between predicted and experimental disruption scores • target: 0.85 • range: -1-1

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

  • Build a domain-specific Knowledge Graph Index with LlamaIndex

  • Integrate Hugging Face embeddings and ESM protein language models

  • Calculate protein interaction perturbance scores for missense mutations

  • Query multi-modal structural proteomic datasets using structured Cypher/SPARQL semantics

Resources and assets

Reference links and supporting material

Dataset notes

Experimentally validated dataset of 300 autism-associated gene mutations with quantitative protein-protein interaction loss/gain measurements.

Challenge setup

Evaluates correlation between predicted interactome disruption scores and experimental binding assay data.

Challenge input

JSON with protein_a, protein_b, mutation (e.g. 'P102L'), and wildtype structures.

LlamaIndex

Framework required for combining structural knowledge graph indexing with LLM reasoning.

Hugging Face

Provides pre-trained ESM protein language model embeddings.

Evaluated output

JSON object with calculated_disruption_score (0.0 to 1.0) and supporting network node path.

Evaluation checks
  • Ensure disruption score is bounded between 0.0 and 1.0.
  • Pearson correlation coefficient between predicted and experimental disruption scores • target: 0.85 • range...
Configured evaluation evidence
  • Benchmark: Proteomic Mutation Perturbation Benchmark
  • Pearson Correlation target: 0.85
  • 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
  • Hugging Face · Required
  • Exa · Optional
  • Llama Index · Optional

Evaluation contract

  • score_range_check · Weight 1
  • pearson_correlation · Weight 1

Metrics

  • pearson_correlation · Target 0.85

Benchmarks

  • Proteomic Mutation Perturbation Benchmark
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with protein_a, protein_b, mutation (e.g. 'P102L'), and wildtype structures.

    Output: JSON object with calculated_disruption_score (0.0 to 1.0) and supporting network node path.

    Execution harness: Python

Evaluation tools

  • LlamaIndex · Graph RAG Framework
  • Hugging Face · Model Model Repository

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Interactome Mutation Impact Analyzer using LlamaIndex and Hugging Face