In order to increase patient care in the time of individualized medicine, it is important to understand the intricate links that exist between medications, genes, and diseases. The aim of this research project is to exploit the PharmGKB database in order to forecast various interactions. This research seeks to determine the relationships that exist between diseases, genes, and drugs, and such a task will is performed with the aid of machine learning techniques and evidence-based practice. The collection also includes many other types of relations such as gene-drug relationships, disease-drug, drug-haplotype, and gene-disease relationships. Several sources, such as locus VIPs, variant clinical, FDA medication prescriptions, and doses, and PharmGKB pathways help to establish these links. While associations have been made in VariantAnnotations or Clinical Annotations, it is often the case that a specific genetic variation or haplotype and the illness phenotype are not causal. This has to do with the fact that an association does not have to draw a direct link to diseases but rather explain how treatments work on patients with those ailments. Relationships among objects can be predicted, for example, using machine learning methods such as clustering or classification. Clustering algorithms sort items into groups based on similarities, while classification models determine the status of a relation between objects. The results of this research provide new important insights in the area of the interrelations of medicines and genes and diseases that takes a step forward in personalized medicine. This allows for the making of more informed decisions regarding the interpretation of treatment outcomes and suitable individualized strategies for therapy. It appears that pharmacogenomics and personalized medicine research are heading somewhere, as machine learning and evidence-based analysis are there to assist.