AI-based multimodal integration of genomics and electronic health records

The widespread adoption of electronic health records (EHRs), which capture patient-specific longitudinal information on diagnoses, laboratory tests, clinical procedures, and outcomes, has created unprecedented opportunities to study diseases at scale. Integrating EHR data with genomic information offers novel ways to understand disease heterogeneity, identify biomarkers and therapeutic targets, and predict disease risk to improve clinical decision-making at scale. Recent methodological advances in machine learning (ML) and artificial intelligence (AI) can handle data with high dimensionality, high levels of noise, and irregular temporality better than traditional statistical approaches. Progress in EHR-linked biobank development, data standardization pipelines, and architectures for modelling and implementation have accelerated the advancement of the field and warrant an assessment of current capabilities and limitations. This Review highlights AI and ML frameworks for integrating genomic, multi-omics, and EHR data, and discusses how these approaches are reshaping genomics research as well as clinical practice.