Project Document: Using EVO2 for Predicting Impacts of Non-Coding Regions Variants

Purpose

The purpose of this project is to utilize EVO2, a DNA foundation model developed by the Arc Institute, to predict the impacts of certain genetic variants, with a specific focus on non-coding regions and their influence on protein expression and phenotype related to Ataxia. The goal is to determine whether any of the numerous variants in non-coding regions could contribute to Ataxia phenotypes.

Scope

This project will focus on:

• Exploring the capabilities of EVO2 in analyzing non-coding genetic variants and their potential influence on Ataxia phenotypes.
• Developing a workflow to process variant data and predict potential impacts using EVO2.
• Utilizing and adapting existing test data, derived from the Snakemake Workflows for DNA sequencing (https://github.com/snakemake-workflows/dna-seq-varlociraptor) and vembrane output files (https://github.com/vembrane/vembrane), to facilitate this analysis.

Objectives

• Gain familiarity with the EVO2 model and its applications.
• Develop a processing pipeline that can accept TSV files with variant data.
• Evaluate the impact of specific variants on protein expression, focusing on non-coding regions.

Deliverables

• A fully functional pipeline that accepts TSV files of variants, processes them, runs them through the EVO2 model, and outputs results.
• A report detailing the methods used, challenges faced, and results obtained.
• Documentation for future users to replicate or expand the analysis.

Timeline

• Project initiation and background research: TBD
• Initial setup and exploration of EVO2: TBD
• Workflow development and testing: TBD
• Analysis and reporting: TBD
• Final presentation: TBD

Methodology

• Utilize Python and an appropriate workflow management tool compatible with Slurm.
• Use test datasets from the WES_Analysis_11340_9740 repository as initial input.
• Implement and refine a pipeline for processing and analyzing variants using EVO2.

Resources

• EVO2 model details from Arc Institute: Arc Institute EVO2
• Test data: WES Analysis Repository
• Workflow management tools: TBD (consider exploring alternatives to Cookiecutter Data Science for Python and Slurm compatibility)

Evaluation Criteria

• Accuracy and reliability of the workflow and predictions.
• Clarity and completeness of documentation and reporting.
• Innovation in workflow design and problem-solving approaches.

Background Information

EVO2 is a genome modeling and design framework capable of analyzing DNA sequences across various domains of life, particularly valuable in studying non-coding regions.

Support and Guidance

• Regular check-ins and feedback sessions: TBD
• Access to shared resources and guidance from [Your Name]: TBD

Challenges and Constraints

• The need to adapt to a currently unfamiliar model (EVO2).
• Ensuring compatibility with cluster resources and data formats.

Communication

• Preferred channel for communication: TBD
• Frequency of updates and meetings: TBD

Risk Management

• Identify potential technical obstacles early and seek guidance as needed.
• Regularly back up work and maintain version control.