SUN-476 THE HUMAN CELL ATLAS AND ARTIFICIAL INTELLIGENCE: KIDNEY MEDICINE IN THE DIGITAL AGE AND BEYOND

Abstract

The Human Cell Atlas (HCA) Project - Single-cell RNA sequencing (scRNA-seq) is now a clinically viable technology that can distinguish between cells that are morphologically identical but genetically and functionally distinct, based on active gene transcription. The exponential development of sequencing technologies in the past decade allows us to sequence millions of cells at a time, opening the door for assembling whole-organism genetic libraries using the single-cell approach. This is the target of large-scale initiatives like the multiple organ-specific components of the Human Cell Atlas (HCA). scRNA-seq has already started to augment our understanding of healthy and pathological renal cells. Following the trends observed in analogous systems, in which scRNA-seq doubled the number of known cells, we predict that renal cells will experience a two-fold increase from 26 to 52 classified cell subtypes. The proposal of a new Banff Classification of Tissue Engineering Pathology (TEP) based on the HCA will have a major impact on kidney medicine diagnostics and patient care, making both of these initiatives central in the practice of transplant physicians and transplant pathologists (The Human Cell Atlas Project by The Numbers - Relationship to the Banff Classification. (2018). American Journal of Transplantation, 18(7), 1830.) We examined the latest studies, analyzing various applications of new medical technology on kidney medicine. Exponentially improving technologies like scRNA-seq, and the HCA in particular, stimulate the need for a new Banff Classification of TEP. This new classification was suggested in 2011, added permanently to the Banff consortium in 2017, and now has concrete plans to be operational by 2025. Specific methods on incorporating machine learning and HCA findings will be proposed and discussed at upcoming Banff Foundation meetings, starting with the 2019 meeting which was held in Pittsburgh, and continuing in the 2021 Meeting (Banff, Canada), the 2023 meeting (Paris, France), and the 2025 meeting (Leuven, The Netherlands). The mass collection of data from increasingly high-yield technologies like scRNA-seq requires powerful machine learning algorithms for data mining. A new wave of research has emerged, combining the skillsets of medical experts and computer scientists, using methods such as digital pathology, gamification, and crowd-sourcing. With the help of the HCA and other similar projects, we can begin to understand how kidney medicine will evolve in the Digital Age and beyond. The combination of high-throughput medical technology and machine learning has promising outcomes, some of which are already a reality. scRNA-seq studies have identified novel cells, linked specific altered gene expression to human diseases, and mapped disease progression to help create precision therapeutics. With the help of powerful data mining algorithms, we can improve prognostication and prediction (such as allograft survival outcomes) and standardize histopathology analysis with computer-assisted diagnosis Together, the HCA and the new Banff Classification of TEP will increase the success of transplantation of bioengineered organs and decrease the number of grafts lost. Here we examine the literature to outline some of the major strides taken in kidney pathology and transplantation through the use of scRNA-seq and artificial intelligence.