The global healthcare burden of Clostridioides Difficile (C. difficile) has worsened, given the increased prevalence of hypervirulent strains across community and hospital settings. To manage transmission, the gold standard of strain typing, PCR-ribotyping, is conducted but requires high effort and resources. As whole genome sequencing data is informative of PCR-ribotype and widely publicly available, this project proposes an adoptable and interpretable C. difficile PCR-ribotype prediction machine learning model, trained on whole genome sequences. Sensitive personal data is not used in this project.