A Bioinformatic Toolkit for Single-Cell mRNA Analysis

The recent technological developments in the field of single-cell RNA-Seq enable us to assay the transcriptome of up to a million single cells in parallel. However, the analyses of such big datasets present a major challenge. During the last decade, a wide variety of strategies have been proposed covering different steps of the analysis. Here, we introduce a selection of computational tools to provide an overview of a generic analysis pipeline.The first step of every scRNA-Seq experiment is proper study design, which does not require sophisticated experimental or informatics skills but is nonetheless presumably the most important step. The quality of the resulting data strictly depends on the proper planning of the experiment, including the selection of the most suitable technology for the biological question of interest as well as an elaborated study design to minimize the influence of confounding factors. Once the experiment has been conducted, the raw sequencing data needs to be processed to extract the gene expression information for each cell. This task comprises quality assessment of the sequenced reads, alignment against a reference genome, demultiplexing of the cell barcodes, and quantification of the reads/transcripts per gene. As any other transcriptomics technology, single-cell mRNA-Seq requires data normalization to assure sample-to-sample, here cell-to-cell, comparability and the consideration of confounding factors.Once gene expression values have been extracted from the reads and normalized, the researcher has the agony of choosing between a plethora of analysis approaches to investigate diverse aspects of the single-cell transcriptomes, such as dimensionality reduction and clustering to explore cellular heterogeneity or trajectory analysis to model differentiation processes.In this chapter, we present a wrap-up of the abovementioned steps to conduct single-cell RNA-Seq analyses and present a selection of existing tools.