Development and validation of a multi-class molecular model for differential diagnosis of renal allograft rejection, built upon gene expression data from the B-HOT gene panel in a large and diverse cohort of diagnostic FFPE renal biopsies. Overall design: In this study, gene expression profiling analysis of 770 genes, including 12 housekeeping genes and 758 target genes, was performed for the FFPE renal biopsy tissue of totally 1395 samples, retrieved from clinical pathology archive, with the B-HOT gene panel assays on NanoString nCounter® platform. All 1395 samples are divided into 4 different groups based on the clincial histology diagnosis related to kidney transplant rejection status: antibody-mediated rejection, T cell-mediated rejection, mixed rejection, and no rejection controls.