Introduction: The high dimensional and heterogeneous “phenomics” data captured in modern Dairy farms pose data integration, harmonization, modelling, and analytical challenges. Most modern dairy farms in EU also collect genotype data from either genomic arrays or genome sequencing and sometimes other omics – e.g. epigenomics and metabolomics. Phenomics data may arise from robotic feeding systems, milk, fertility and health recording systems, animal behavioural and activity logger system, milk components / biochemical profiling, environmental monitoring etc. In the new era of phenomics and multiomics, precision dairy cattle breeding and management for Performance, Health, Resilience, Efficiency, and Fertility (PHREF) traits are not fully optimised for phenomics and genomics / multiomics datasets and needs improvement. Methodologies: The iMAP-Dairy use case project aims to develop a framework to integrate, model, and analyse phenomics and multiomics data from novel technologies used in modern dairy farms. It applies Artificial Intelligence -Machine Learning (AI-ML) methods as well as statistical genetics and bioinformatics methods to process, model and analyse PHREF data. Denmark Cattle Research Centre (DKC) of the Aarhus University provides such datasets collected over 10-year period. We will use Deep learning (DL) methods such as Fully Connected Neural Networks (FCNN), Convolutional Neural Networks (CNNs), Graph Neural Networks (GNNs) for prediction while supervised ML methods will include Lasso/ElasticNet, Random Forest and XGBoost. Output: Harmonized data integration workflows for phenomics, and multi-omics data. AI-ML based analytical protocols to predict future PHREF traits. Through supervised ML and GWAS analyses, identified key genetic variants/genes, biological pathways, environmental, and management factors influencing PHREF traits. Impact: The iMAP-Dairy project will enable predictive, data-driven decision-making for more sustainable, climate smart, and welfare-oriented dairy production systems and precision dairy farming. What I can offer: I can contribute to collaborative research projects within AgDATA partnership via internal or external calls. Our area of expertise includes G2P data integration and analyses involving phenomics and genomics/multiomics data types and use of AI/ML and Bioinformatics approaches to analyse them. What I need: To strengthen the modelling framework and improve prediction accuracy, we need access to external, independent datasets for validation from other partners or countries.

Source: Annual AgData Use Case Summit 2026 booklet.