A critical component in predicting how breast cancers are likely to behave and whether treatments such as chemotherapy are required, is the tumour grade. Currently pathologists assess grade using microscopic assessment of a patient's tumour on a glass slide, and the result can vary between specialists (i.e. subjectivity). Pathologists in the near future will replace looking at slides with looking at digital images of the cancer on a monitor. This will also allow the application of artificial intelligence algorithms (computational pathology) to assist the pathologist in making more accurate and less opinion-based (i.e. more objective) decisions. This will lead to more robust cancer assessments to guide tailored treatments and improve outcomes. This application aims to foster computational pathology capability and expertise by using this approach to explore the nature of the relationship between the microscopic appearance of breast cancer (nuclear pleomorphism) and the associated molecular changes, informing future project planning.