The GenomeComputer Company
Back to tutorials

Tutorial

Understanding your .genome bundle

Learn what your .genome bundle is, how it differs from raw sequencing files, and how to use it as the starting point for AI-assisted genome exploration.

6 minAny AI tool

What you will learn:

You will be able to open a .genome bundle, understand what each file is for, and choose the right starting point for a question.

A .genome bundle is the practical layer between raw sequencing data and useful AI exploration. Raw files are built for sequencing machines and bioinformatics pipelines. A .genome bundle is built for you, your AI tools, and any workflow you want to run around your own genome.

The simplest way to think about it is this: FASTQ and gVCF preserve what was sequenced, while .genome organizes what can be explored. It packages genome data, annotations, source references, evidence context, and reusable prompts into a format that AI tools can inspect without wasting context on unnecessary file noise.

This matters because the first job of an AI tool is not to give you a sweeping interpretation. The first job is to understand what data it has. A .genome bundle gives the model a structured map: what files exist, what they contain, which annotations are available, and where evidence came from.

What is usually inside

  • AI-readable genome summaries that help a model orient itself before answering.
  • Variant and annotation files that connect genes, variants, traits, sources, and evidence levels.
  • Source or citation fields that make it easier to trace claims back to research or databases.
  • Prompt files or examples that show the kinds of questions the bundle is designed to support.
  • Technical files or references that preserve deeper sequencing context when available.

What the bundle is for

  • Starting a conversation with your genome in Codex, Claude Code, or another project-based AI tool.
  • Asking narrow questions about genes, variants, traits, pathways, or research papers.
  • Separating what is directly present in your file from what is inferred from annotations or broader literature.
  • Reusing the same exploration workflow as research changes over time.
  • Keeping your genome in a format that you can download, self-host, archive, or inspect locally.

First pass workflow

  1. Open a new AI conversation or local workspace.
  2. Upload or open the .genome bundle.
  3. Ask the model to inspect the files before answering any biological question.
  4. Ask it to create a file map with purpose, likely contents, and best use case.
  5. Ask which files are best for broad exploration, variant lookup, source review, and technical review.
  6. Save the file map. This becomes your reference point for every future exploration.

Map the bundle

I have opened my .genome bundle. Before answering any genetics questions, inspect the available files and create a map of the bundle. For each file, explain what it appears to contain, what it is useful for, and whether it is best for broad exploration, variant lookup, source review, or deeper technical review.

Choose the right file

I want to explore [topic]. Based on the files in my .genome bundle, tell me which file or files you should inspect first. Explain why those files are relevant, what evidence they can provide, and what they cannot answer on their own.

What a good answer looks like

  • It mentions the actual files in your uploaded bundle, not a generic list of possible genome files.
  • It separates file purpose from interpretation.
  • It tells you which file to start with for a broad question and which file to use for a deeper check.
  • It is honest about missing files, unreadable files, or evidence that is not present.
  • It gives you a next question to ask rather than pretending one pass is the whole analysis.

Once you have this map, your .genome becomes much easier to use. You can ask about a gene, variant, pathway, paper, or trait and require the model to ground its answer in the bundle rather than answering from memory.