OBJECTIVES: Existing solid organ transplant models are based upon organ availability, donor compatibility and general measures of severity but they lack the ability to fully capture the personalized context of treatments that certain instances can avoid the need for transplantation and re-transplantation. In this study, we propose supervised and unsupervised Bayesian Network Models to predict the most personalized pathways that minimize the disease progression conditional upon patient needs and access to the most suitable technology. Furthermore, in addition to prediction, we will also test how well patients would perform if given access to technology that can slow the progression of illness and assess the impact of these interventions on likely rate of transplant and downstream costs. METHODS: Using the United Network for Organ Sharing (UNOS) National Organ Procurement and Transplantation Network (OPTN) dataset from 2000-2015, we developed a Bayesian Network Model to estimate the joint probability distribution bilaterally, between donor and recipient, to predict transplant survival rates conditional upon defined biological, clinical and treatment variables. RESULTS: Preliminary results in kidney transplantation show that increases in survival rates are correlated not only with host characteristics or donor characteristics, but also treatment characteristics. A k-fold validation of the Bayesian Network Model shows more than 60% predictive power for survival rates. Such models can be made available graphically to transplant teams to assist in optimizing donor characteristics and treatments to the precise needs of kidney patients. CONCLUSIONS: The precision medicine movement of today not only requires access to life saving technologies, but also access to pre-emptive information on donors, hosts and treatment options that can be personalized to unique biological and behavioral variables.