Developing tools and standards for integration of multidimensional HCA data

Technical advances in RNA-Seq from individual cells make it possible to profile hundreds of thousands of single cells and to build a Human Cell Atlas (HCA). However, current large-scale efforts to map the cellular composition of tissues and organs largely rely on the analysis of cells after dissociation, a process potentially distorting gene expression profiles and biasing sample composition. Thus, this project aims to comprehensively benchmark current single cell approaches to assess the reproducibility, integrity and predictive value of state-of-the-art data production tools. Multiple transcriptomic and epigenomic datasets from different donors and sampling sites will be integrated to evaluate performance, complementarity and replicability. Specifically, datasets for ten human kidney samples will be generated through single-cell and single-nucleus RNA sequencing methods; complemented with epigenomic (single-cell ATAC-seq) and multi-omics (single-cell NMT-seq) technologies. A validation phase with spatial RNA (MERFISH) and protein (CODEX) information will be followed by data integration and power analysis to identify technical biases, complementarities, and redundancies of the methods. Thus, the main goal is to align the experience, methodologies and computational pipelines available in the group to provide a framework of technologies, quality metrics, guidelines, and new computational tools that will serve as a compass for the HCA’s generation of reproducible and high-quality tissue atlases.