The聽CoRAL聽(Consensus Reference-based Automated Labeling) pipeline,聽is an open-source tool that uses a weighted consensus strategy to predict cell type identity from single-cell brain reference atlases. Developed in the laboratory of聽Claudia Kleinman, PhD聽this automated platform currently incorporates predictions from 19 different machine learning-based tools, outperforming individual tools in a cell annotation task, and allowing to obtain, for disease samples, a quantitative estimation of pathological deviations from healthy.聽
CoRAL聽is continually updated by members of the聽聽to incorporate newly available tools and information. The most recent updates integrated a weighted consensus to the analysis to prioritize predictions based on tool performance. This Open Science tool, prioritizing easy installation and running, is available for all researchers looking to analyze single-cell
data on the聽Github聽page of the Kleinman Lab.聽
