A clustering method to study the loss of kidney function following kidney transplantation

Clinical outcomes from the Assessing Donor-derived cell-free DNA Monitoring Insights of kidney Allografts with Longitudinal surveillance (ADMIRAL) study

“…This was used to calculate the serial delta change between dd-cfDNA results associated with pathology using the methods outlined by Lund et al 13 Analytical variation was defined as 2.7%, intrapatient variation ¼ 61%, and the index of individuality ¼ 0.23%. 12 K-means clustering 14 was used; distinct clusters representing time points allowed the formation of time horizons from 0 to 3 years post-transplant. The machine learning algorithm partitioned data into monthly clusters that were predetermined by minimizing the sum of squared distance using key features such as ethnicity, sex, age at transplant, evidence of BK virus infection, dd-cfDNA score, presence of DSAs, allograft rejection, and creatinine.…”