Diagnosis of T-Cell-Mediated Kidney Rejection by Biopsy-Based Proteomics and Machine Learning

PurposeThis study is aimed at developing a clinic-friendly proteomics protocol and a machine learning (ML)-based molecular diagnostic test for T-cell-mediated rejection (TCMR) using formalin-fixed, paraffin-embedded (FFPE) biopsies. Experimental designBased on the procedures we reported for proteomic profiling of FFPE biopsies using Tandem Mass Tag (TMT)-based technology, a label-free-based quantitative proteomics protocol was developed as a more clinical-practical and cost-efficient molecular diagnostic test for renal transplant injection. This new protocol was applied to a set of FFPE biopsies from renal allograft injury patients and normal controls, including 5 TCMR, 5 polyomavirus BK nephropathy (BKPyVN) and 5 stable graft function (STA). Three different machine learning algorithms, linear discriminant analysis (LDA), support vector machine (SVM) and random forests (RF), were tested to build a prediction model for TCMR. ResultsAbout 750-1250 proteins were identified and quantified in each sample with high confidence using the label-free-based proteomics protocol. 178, 450 and 281 proteins were defined as differential expression (DE) proteins for TCMR vs STA, BKPyVN vs STA and TCMR vs BKPyVN, respectively. By comparing the quantitative data from the TMT- and label-free-based proteomics profiling, a classifier panel comprised of 234 DE proteins commonly quantified by two methods was generated to test different ML algorithms. Leave-one-out cross-validation result suggested that the RF-based model achieved the best prediction power for TCMR at both proteome and transcriptome level. Conclusions and clinical relevanceProteomics profiling of FFPE biopsies using a platform integrated of label-free quantitative proteomics with ML-based predictive model can help to discover biomarker panels and provide clinical molecular diagnostic tests to enhance biopsy interpretation for renal allograft rejection. Clinical RelevanceThis study is to develop a molecular diagnostic test for kidney rejection. An easy-to-use and cost-efficient protocol using label-free quantitative strategy was developed to profile proteome of FFPE biopsies from kidney allografts. A list of 234 DE identified from TCMR, BKPyVN and STA was generated as a classifier panel for these different phenotypes. This classifier panel was subjected to the optimized ML model, achieving high accuracy among both positive and negative control. This proof-of-principle study demonstrated the clinical feasibility of implementation of molecular diagnostic tests integrated of label-free-based quantitative proteomics and ML-derived disease predictive models to enhance biopsy interpretation for kidney transplantation patients. More accurate and specific molecular tests can lead to more effective treatment, prolong graft life, and improve the quality of life for patients with chronic kidney failure.