Aim 1 : We will collaboratively derive best practices, metrics, and benchmarks for integrating heterogeneous scRNA-seq datasets.
Accompanied by new metrics and quantitative benchmarks, this aim will facilitate the development and the comparison of alternative methods for scRNA-seq data integration.
Aim 2 : To demonstrate the power of this approach, we will integrate all available transcriptomic datasets profiling the human nervous system, including sparse, deep, cytoplasmic, and nuclear single cell transcriptomics.
This will provide a clear roadmap for HCA to construct a coherent atlas of cell states from diverse community-generated datasets, demonstrating the strengths of integrated analysis for cellular phenotyping.