Abstract:
|
Single-cell RNA sequencing enables the profiling of gene expression at the single-cell level for a large sample of cells. This allows researchers to address many important biological questions, such as the investigation of rare cell types and the discovery of novel subpopulations of cells from a heterogeneous population. As in other high-throughput assays, batch effects can have a large impact on the data; hence a careful experimental design is essential to ensure that the question of interest can be investigated. Even with an optimal design, unknown sources of unwanted variation can still affect the data, and normalization is an essential step prior to any inferential procedure. I will discuss some design, quality control (QC), and exploratory data analysis issues of single-cell experiments, using data from the Fluidigm C1 platform, highlighting how simple QC metrics can be used for sample filtering and normalization; I will illustrate how unwanted technical effects influence clustering and differential expression analyses and how they can be taken into account in a general statistical model.
|