Pacific Analytics at the GA4GH 13th Plenary: Advancing Responsible Genomic Data Sharing and AI Innovation

The GA4GH 13th Plenary and Connect working sessions — co-hosted by SciLifeLab in Uppsala, Sweden — brought together nearly 300 in-person attendees and more than 300 virtual participants for a week of collaboration, technical progress, and strategic alignment. The annual Plenary continues to be a critical forum uniting researchers, healthcare organisations, government representatives, technologists, and industry partners to shape the future of responsible genomic and health data sharing.

Pacific Analytics was represented by Boris Guennewig, Alex Kanitz, Leisa Rumjahn, Anurag Gupta, and Javed Habib contributing across group discussions, whiteboard sessions, and the poster showcase to share our work in secure, standards-aligned platforms and our vision for AI-enabled scientific discovery.

This year's meeting centered on four core themes that resonated strongly with our work:

  • the growing value of genomic medicine,
  • the rapid rise of artificial intelligence in genomics,
  • the importance of standards and global collaboration, and
  • the central role of trust within the health data ecosystem.

Agentic Science and the "Perfect Circle" of AI-Driven Discovery

A major highlight was the establishment of the new GA4GH Artificial Intelligence Work Stream — a recognition of how rapidly AI is reshaping genomic science and data analysis.

As part of this theme, our CEO Boris Guennewig presented "A Perfect Circle: The Scientific Roundtrip Fueled by GA4GH & Agentic AI". He outlined how Pacific Analytics has spent recent years building secure, modular, GA4GH-aligned platforms designed to support real-world consortium research:

  • cloud-agnostic, avoiding vendor lock-in and allowing partners to integrate their own DRS, WES, TES, or other GA4GH components;
  • highly secure, developed on hardened images and ISO-compliant infrastructure;
  • rooted in open-source technologies, including Keycloak, OPA, Prometheus, Trino, and others;
  • auditable and interoperable, enabling real-world federated analysis across organisations.

Boris also introduced our Agentic AI framework, which supports a fully traceable "scientific roundtrip" — from hypothesis generation to dataset identification, workflow execution, and analysis. Key capabilities include large-scale literature ingestion, hypothesis generation informed by curated collections of genomics and biomedical literature, ELO-style feasibility scoring, GA4GH TRS-driven tool selection, WES-aligned workflow execution, and code-interpreter-powered analytics. Researchers remain fully in the loop today, with increasing opportunities for semi-automated scientific exploration under appropriate governance.

GA4GH in the Real World — Interoperability at Scale

Anurag Gupta and Javed Habib presented Pacific Analytics' real-world application of GA4GH standards as part of the session "GA4GH in the Real World." They were joined by CTO Alex Kanitz, who contributed to a lively and highly interactive Q&A session.

The audience was especially interested in how Pacific Analytics operationalises GA4GH APIs in production environments — spanning authentication and authorisation flows, federated compute models, multi-cloud deployment, workflow orchestration, and integrating WES, TES and DRS components into functioning research ecosystems.

TRE Open Suite — Building Trustworthy, Federated TREs

Another key contribution came from Alex Kanitz, who co-led the session and is an active contributor to the GIF TRE Open Suite project — an ambitious, multi-year initiative to deliver a GA4GH-native framework and SDK for Trusted Research Environments (TREs) that enable secure, privacy-preserving computation across institutions, clouds, and jurisdictions.

The session featured a world cafe setup to engage additional stakeholders from academia and industry. The following topics were discussed:

(1) The project's modular, plug-in architecture, which is designed to support a range of trust-enhancing technologies, including Crypt4GH encryption, Homomorphic Encryption, Differential Privacy, and attested communication.

(2) Alignment with existing and emerging TRE governance frameworks, such as the Five Safes and the EOSC-ENTRUST initiative, recognising that effective TREs require technical and policy foundations to work hand in hand.

(3) The modeling, establishment and continuous verification of complex, cross-organisational trust relationships in federated networks, which is a critical requirement for ensuring that data privacy is preserved at all times.

Throughout the discussion, federation, transparency, and interoperability were identified as core design principles by participants, alongside a pragmatic focus on how TREs can be deployed in complex real-world environments with diverse regulatory, policy, and compliance constraints.

For Pacific Analytics, this work resonates deeply with our own platform designs. Our multi-cloud, GA4GH-aligned, audit-ready architecture is already addressing many of the same challenges, and active participation in the TRE Open Suite initiative further strengthens our role in shaping the next generation of secure and federated research infrastructures.

Looking Ahead

The 13th Plenary and Connect working sessions reinforced the momentum across the GA4GH community: AI is transforming discovery, federated analysis is becoming the norm, and trust — grounded in standards, governance, and transparent systems — is foundational to the future of genomic and health data sharing.

Pacific Analytics is proud to contribute to this work and to help shape the next era of responsible, globally connected biomedical research.