Predict the Future

Overview

Health Hub EDIH, together with GeNeDis and OnMune, runs a one-day hackathon on multi-omics data integration, knowledge mining, and applied AI in translational biomedicine. It builds on an earlier proof of concept and targets the next step: re-runnable data workflows, clean documentation, and technical foundations for a future platform.

The challenge is anchored on two indications — Non-Small Cell Lung Cancer (NSCLC) and Rheumatoid Arthritis (RA). Teams may address either indication, or both; a structure that serves both is encouraged but not required. Selected teams present their work at the SNFCC during GeNeDis 2026, before an international scientific audience.

Participation is hybrid You can take part in person at the SNFCC or remotely online.

Prize

As the hackathon award, the winning team is offered free publication of their work as a Book Chapter in the conference's proceedings volume, in which part of the GeNeDis 2026 proceedings is published, plus a 3-month internship at OnMune. Selected teams also present their work at the SNFCC during GeNeDis 2026, before an international scientific audience.

Task

Build a reproducible pipeline, or a focused knowledge-mining model, that integrates multi-omics data for the chosen scope, surfaces meaningful patterns, and retrieves relevant biological information. The priority is clean data organisation, repeatable steps, and documentation another team can follow.

Where a supervised target is used, define labels explicitly, check for data leakage, and compare against simple baselines. Results are treated as exploratory signals for further study, not clinical predictions.

Choose one scope:

• NSCLC — tumour cohorts, preferably lung adenocarcinoma and squamous (LUAD / LUSC).
• RA — rheumatoid arthritis, with synovial tissue as the primary axis.
• Both — a cross-indication structure; encouraged and viewed positively, but not required.

Data

Work from open multi-omics resources, or from datasets your team already holds legitimate access to. Do not rely on new controlled-access approvals being granted during the event. Each team documents its sources, selection criteria, labels, and how processed files were produced.

Evaluation

Entries are not ranked by a single numeric score. A committee evaluates each submission across four dimensions:

1. Data organisation. How cleanly the multi-omics base is built for the chosen scope, and whether sources, accession lists, labels, metadata, access terms, and processed files are documented.
2. Method. Whether AI/ML or information retrieval is used appropriately — clear labels, no data leakage, and a baseline comparison wherever a supervised target exists.
3. Biological grounding. Whether results are explained biologically, and whether candidate biomarkers, pathways, or signatures are scientifically justified for the indication, or across both.
4. Reproducibility. Whether the workflow re-runs, extends to new datasets or disease contexts, and reconstructs its core files from the declared steps.

Technical delivery is weighed alongside the science: install instructions, a clean README, a documented data structure, a source/accession and provenance manifest, a QC or model report, versioned outputs, and a clear distinction between ready, exploratory, and unavailable results.

Eligibility

Open to anyone who wants hands-on work across biomedical data, machine learning, software, and data organisation:

• Undergraduate and postgraduate students (Bioinformatics, Data Science, Biology, Medicine, Informatics)
• Researchers and PhD candidates
• ICT and healthcare professionals
• Software developers and data analysts

Enter individually or as a team. Participation is free and hybrid.Join on-site at the SNFCC or online.

Important dates

Submission

1. Register to receive the challenge brief, indicative NSCLC/RA data sources, and the submission template.
2. Build a reproducible workflow for NSCLC, RA, or both that integrates multi-omics data and produces documented outputs via ML/AI or information retrieval.
3. Submit your deliverables by the deadline (to be announced) to aris.vrahatis@ionio.gr. Late submissions are not accepted.
4. Shortlisted entries are presented at the Health Hub event, on a date to be announced.

Deliverables

• A short README
• A source / accession list with documentation
• The code or pipeline
• Processed feature tables
• Key outputs, or a representative sample
• A QC / model report
• Clearly stated limitations

Data & ethics

All data must be secondary, open or properly licensed, and fully anonymised. Submissions must contain no passwords, private keys, or identifiable personal or clinical data. Data use is governed by GDPR, NIS2, the DSA, and the European Health Data Space (EHDS), now in force and applying progressively.

Organizers & partners

Health Hub EDIHEuropean Digital Innovation Hub.
healthhub-edih.eu

GeNeDis 20267th Genomics, Neuroscience, Therapeutics and Data Innovation Summit

OnMune
onmune.com

FAQ

Is there a participation fee?

No. Participation is free, for individuals and teams.

Is participation on-site or online?

It is hybrid. You can take part in person at the SNFCC or remotely online.

Is there a prize?

Yes. As the award, the winning team receives free publication of their work as a Book Chapter in the GeNeDis 2026 proceedings.

Can I enter on my own?

Yes. Enter individually or as a team.

Which indications can I choose?

NSCLC, RA, or both. A structure serving both is encouraged but not required.

Do I need controlled-access data?

No. Use open data, or data you already hold approved access to. You cannot rely on new controlled-access approvals being granted during the event.

What do I submit?

A README, a source/accession list, your code or pipeline, processed feature tables, key outputs, a QC/model report, and stated limitations.

When and where?

21 November 2026 at the Stavros Niarchos Foundation Cultural Center, Athens, within GeNeDis 2026 — on-site or online. The exact time and the submission deadline will be announced.