“…The comparison of predictive performances between our classifier and classifiers in the literature underlines its relevance to pre-kidney transplant risk assessment (Supplemental Table 5). Its accuracy of 82.7% is i) one of the highest among all donor-independent risk assessment models (19,25,30,34,57), ii) comparable to any AR models (19,20,30,34,35,57,58,21,22,(24)(25)(26)(27)(28)(29) and iii) comparable to any SAB data based models for ABMR (22,(24)(25)(26)(27)(28). Furthermore, our classifier is based on SAB, an established diagnostics laboratory tool, thereby facilitating its further use for ACR risk assessment.…”