Genomic medicine, which provides diagnosis and treatment decisions based on genomic variations, has been implemented in clinical practice and become more accessible. The clinical interpretation of genomic variations detected through genome analysis is crucial in genomic medicine. Clinical interpretation often utilizes genomic databases and effects predictions by machine learning. However, it has been noted that the genomic data accumulated thus far exhibit biases toward certain ethnic groups, leading to efforts to enhance genomic data diversity across various regions. While diverse and highly accurate machine learning methods have been developed for coding region variants, prediction methods for noncoding region variants still need to be improved in accuracy and applicable region. This chapter outlines the basic flow of genomic medicine and how detected variants are interpreted, introducing disease variant databases to advance genomic medicine and machine learning methods for predicting the effect of noncoding region variants.