AutoMRAI: A Multi-Omics Causal Inference Platform Using Structural Equation Modelling

The integration of causal inference, artificial intelligence (AI), and multi-omics data represents a transformative frontier for unravelling the complex mechanisms underlying health and disease. Traditional observational epidemiology is limited by confounding and reverse causation, while statistical frameworks such as Mendelian randomization (MR) and structural equation modelling (SEM) enable more robust causal inference. Recent advances in AI and machine learning have further expanded these capabilities, providing scalable solutions for biomarker discovery, therapeutic target identification, and precision medicine. Here, we introduce AutoMRAI, a unified platform that integrates SEM with multi-omics data analysis in an AI-augmented environment. AutoMRAI enables causal modelling across genomic, epigenomic, transcriptomic, proteomic, metabolomic, microbiome, and clinical layers, supported by directed acyclic graph (DAG) representation, robust statistical modelling, and interactive visualization. We provide a proof-of-concept demonstration using simulated datasets, highlighting the platforms ability to recover known causal pathways across omics layers. AutoMRAI addresses key challenges of reproducibility, scalability, and interpretability in causal inference and establishes a foundation for future integration of MR pipelines, Bayesian networks, and advanced AI workflows. The platform is offering researchers a scalable and accessible resource for multi-omics causal discovery.