Multi-omic factor analyses uncovered cross-compartment complexity of biological processes in kidney transplantation.

Abstract

Understanding the complexity of pathways involved in allograft injuries after solid organ transplantation is essential for precise definitions of rejection subtypes and improved overall outcomes. High throughput technologies and the recently available computational methods make it now possible to address such complex biological questions. Here, we performed a unique compilation of six omics datasets (170,000 variables) from a multi-center study encompassing 131 kidney transplant recipients. Using multi-omics factor analysis (MOFA), we investigated sources of variability in patient blood, urine and their allograft at the epigenetic and transcriptomic levels. Integrating the different omics layers, MOFA delimited eight hidden factors in an unsupervised manner. We identified specific factors that reflect allograft rejection and their multicellular complex immune profiles, complement activation, monocyte crosstalk, or immune modifications associated with induction treatment. These cross-compartment large datasets translated into an understandable biological picture provide a new framework to solve complex biological questions, not unique to transplant medicine.