Validation of systems biology derived molecular markers of renal donor organ status associated with long term allograft function

Donor organ quality affects long term outcome after renal transplantation. A variety of prognostic molecular markers is available, yet their validity often remains undetermined. A network-based molecular model reflecting donor kidney status based on transcriptomics data and molecular features reported in scientific literature to be associated with chronic allograft nephropathy was created. Significantly enriched biological processes were identified and representative markers were selected. An independent kidney pre-implantation transcriptomics dataset of 76 organs was used to predict estimated glomerular filtration rate (eGFR) values twelve months after transplantation using available clinical data and marker expression values. The best-performing regression model solely based on the clinical parameters donor age, donor gender, and recipient gender explained 17% of variance in post-transplant eGFR values. The five molecular markers EGF, CD2BP2, RALBP1, SF3B1, and DDX19B representing key molecular processes of the constructed renal donor organ status molecular model in addition to the clinical parameters significantly improved model performance (p-value = 0.0007) explaining around 33% of the variability of eGFR values twelve months after transplantation. Collectively, molecular markers reflecting donor organ status significantly add to prediction of post-transplant renal function when added to the clinical parameters donor age and gender.