A review in Nature Reviews Genetics examines AI-based multimodal integration of genomics and electronic health records, drawing on cited studies that address the benefits and drawbacks of electronic health record systems, the use of electronic health records to generate phenotypes for research, and the benefits and limitations of genome-wide association studies. The article references research on PheWAS and associations with medical diagnoses and clinical measures across 38,662 individuals from Geisinger, as well as work on translating polygenic risk scores for clinical use by estimating the confidence bounds of risk prediction. It also cites studies on advancements and limitations in polygenic risk score methods for genomic prediction, the application of large language models in medicine, and the use of large language models to identify social determinants of health in electronic health records. Additional cited research covers multimodal machine learning in precision health, multi-omics data integration and interpretation, and deep learning approaches. The review also references a study on large language models improving transferability of electronic health record-based predictions across countries and coding systems, and work on developing phenotype algorithms using electronic medical records and incorporating natural language processing.