About ProtAIomics

Call: HORIZON-MSCA-2024-DN-01-01
Project ID: 101227009
Runtime: 01/01/2026-31/12/2029

Integrating AI in Proteomics: From Data Acquisition to Translational Applications

ProtAIomics is built around a simple but powerful value chain: turn raw mass-spectrometry data into reliable information, convert that information into deep biological knowledge, and channel the resulting insight into actionable solutions for medicine and biotechnology. Achieving this requires three, mutually reinforcing pillars, (1) pushing the frontier of AI-powered data acquisition and analysis, (2) training sixteen doctoral candidates through a rich mix of academic and industrial secondments, and (3) embedding open science, ethical AI and patient engagement at every step. These goals create the foundation for next-generation proteomics technologies and data-driven health innovations.

Read more

Scientific & Technical Objectives

In ProtAIomics experimental and computational work is channelled through three closely linked Research Objectives: Information, Knowledge and Action.

INFORMATION
We will devise new AI architectures that lift today’s data-utilisation ceiling, capturing far more than the ~60 % of spectra typically interpreted, and in doing so maximise the information extracted from every raw MS run.
KNOWLEDGE
Building on richer data, the network will develop integrative machine-learning pipelines that combine proteomics with genomics, transcriptomics and structural biology to reveal proteoform diversity, spatial organisation and interaction networks — effectively turning heterogeneous information into coherent biological knowledge.
ACTION
Finally, ProtAIomics will translate that knowledge into tangible outcomes: graph-based and generative AI models to predict clinical biomarkers, model antibiotic resistance, and engineer bespoke proteins — all while safeguarding trust, interpretability and ethical deployment in healthcare.
Together, these three objectives ensure that every advance at one stage accelerates progress in the next, creating a virtuous cycle where information, knowledge and action continuously reinforce one another to drive proteomics-powered innovation.

ProtAIomics is defined in seven work packages

ProtAIomics unfolds through seven tightly knit Work Packages (WPs)—three scientific engines and four transversal enablers. The scientific heart consists of WP 1 Information, which pioneers AI methods to capture far more signal from raw mass-spectrometry runs and raise data-utilisation rates; WP 2 Knowledge, which merges those richer datasets with multi-omics and structural biology to extract mechanistic insight; and WP 3 Action, which converts that insight into graph-based biomarker discovery, antibiotic-resistance modelling and AI-guided protein or drug design. Surrounding this core are four cross-cutting WPs that keep the science impactful and sustainable: WP 4 steers dissemination, exploitation and public engagement; WP 5 delivers an ambitious doctoral-training and career-development programme; WP 6 handles day-to-day coordination and risk management; and WP 7 safeguards ethics and responsible AI practices.

All seven WPs interlock so that advances in data acquisition flow seamlessly into deeper biological understanding and, ultimately, into real-world biomedical and biotechnological solutions, while training talent, communicating widely and upholding rigorous governance throughout the project lifecycle.