INTEGRAL

The INTEGRAL project aims to transform poultry health management through the development of an integrated, data-driven strategy that combines rapid pathogen detection, AI-based predictive modeling, and microbiome-based interventions. Driven by the urgent global need to ensure food production for a projected population of 9.8 billion by 2050, the project addresses critical challenges in poultry production, including high infection incidence, increased mortality following antibiotic restrictions, and the emergence of antimicrobial-resistant bacteria, offering innovative diagnostic and management solutions tailored to modern intensive farming conditions.

The main objective is to reduce the prevalence and impact of infectious diseases in poultry production while minimizing antibiotic use and increasing productivity. To achieve this, the project pursues several specific goals: (i) establishing a robust baseline of pathogen incidence and creating a curated in silico database of key poultry pathogens with annotated antibiotic resistance markers; (ii) developing rapid long-read sequencing tools capable of providing comprehensive pathogen and resistance profiles within six hours; (iii) designing and validating a high-performance predictive model that integrates microbiome, phenotypic, environmental, and production data using machine learning techniques; and (iv) exploring and deploying microbiome-based solutions through the isolation, characterization, and artificial selection of beneficial microbial consortia to counteract pathogenic infections.

The methodology is organized into five interrelated technical activities that begin with extensive field sampling and data acquisition from diverse poultry farms, followed by molecular characterization in the laboratory using Oxford Nanopore long-read sequencing. These data are then integrated and preprocessed to build predictive models using supervised learning algorithms such as Random Forests, XGBoost, and neural networks, with rigorous cross-validation and hyperparameter tuning to ensure reliability. In parallel, the project investigates innovative microbiome-based interventions through the artificial selection of microbial consortia via in vitro studies, culminating in comparative field trials against conventional practices. A synergistic public-private collaboration is maintained between UVESA, which provides practical operational expertise and farm-level knowledge, and Leitat, which contributes advanced R&D capabilities in diagnostics, data analysis, and technology transfer.

Expected outcomes include a validated rapid diagnostic platform for early pathogen and antibiotic resistance detection, a robust AI-driven predictive model for real-time infection risk forecasting, and innovative, farm-customized microbiome-based management strategies. These results are expected to significantly improve animal health and welfare, reduce economic losses, and enhance the sustainability and competitiveness of poultry production. By bridging the gap between laboratory innovation and field application, the INTEGRAL project not only advances the state of the art in precision livestock farming but also creates a scalable framework with potential applications in other sectors of animal production.

Leitat’s role in the project: