AI-Resource-Landscape for Pathogen Genomics

AI resources relevant to pathogen genomics and public health

Overview

The AI Resource Landscape for Pathogen Genomics is a living repository of artificial intelligence tools, datasets, models, and initiatives relevant to pathogen genomics and public health. The goal is to provide a centralized, structured view of how AI is currently being applied across the field, identify gaps and opportunities for collaboration, and inform future recommendations and benchmarking work.

This repository serves as a foundational reference for the PHA4GE AI Working Group, directly supporting its downstream efforts in:

• AI Applications
• Data and Resource Gaps
• Equitable Access

Background

The project began with a broad data-dump phase, where working group members contributed any AI-related resources they were aware of. These contributions ranged from:

• Published models and academic tools
• Commercial platforms
• Ongoing research efforts and pilot initiatives

All initial entries were captured in a shared Landscape Sheet, which now serves as the raw input for structured analysis, standardization, and long-term maintenance within this repository.

Project Structure and Phases

Phase 1: Resource Compilation

Objective: Collect and clean an initial set of AI resources.

Key activities:

• Open data-dump via shared forms distributed across PHA4GE working groups and partner networks.

• Capture core metadata, including:

  • Source and organization
  • Application area (e.g., pathogen typing, genomic epidemiology)
  • AI/ML method used (with standardized definitions)
  • Accessibility (open-source vs. proprietary)
  • Intended user level (beginner, intermediate, expert)

• Perform light review to:

  • Remove duplicate entries
  • Flag unclear or incomplete submissions

• Develop a shared definitions table drawing on established terminology (e.g., ISO/IEC, AAAI, AMA).

Phase 2: Categorization and Standardization

Objective: Make the landscape queryable and analytically useful.

Key activities:

• Define a standard schema for all entries, such as:

  • Tool or resource type
  • Input data and outputs
  • Use case
  • Target sector (public health, industry, academic)

• Reformat and normalize existing entries to align with the schema.

• Draft an executive summary describing:

  • The scope of the landscape
  • High-level trends and gaps observed

Phase 3: Public Release and Community Engagement

Objective: Enable transparency, reuse, and community contribution.

Key activities:

• Publish the AI Resource Landscape publicly via GitHub and/or the PHA4GE wiki.

• Establish clear contribution guidelines to allow external users to:

  • Add new resources
  • Suggest edits
  • Comment on existing entries

• Promote the resource through PHA4GE communication channels to invite broader public health and research community input.

Phase 4: Maintenance and Expansion

Objective: Keep the resource current and relevant.

Key activities:

• Transition the landscape into a maintained, living resource.

• Schedule quarterly updates to review and incorporate new contributions.

• Encourage ongoing submissions through:

  • Working group meetings
  • Project rotations
  • Periodic calls for input

Repository Contents

This repository will include:

• Structured datasets representing the AI Resource Landscape
• Schema and metadata definitions
• Documentation and summaries of findings
• Contribution guidelines for internal and external contributors

Contributing

Contribution guidelines will be added as the project moves into public release. Contributions are expected to align with the defined schema and include sufficient metadata to support comparison and analysis.