All the functions that take place within a cell are performed through proteins. These proteins are coded within the DNA (Deoxyribonucleic acid) of the cell. A gene is a sequence of DNA that encodes for a particular protein. In order to make the necessary proteins, the transcriptional machinary of a cell makes special copies of the respective genes that can be translated to protein sequences. These special copies are called messenger RNAs (Ribonucleic acids).
The amount of mRNA produced by a specific gene is used as a surrogate marker for quantification of the gene activity. With Current highthroughput sequencing technologies (such as in this case RNAseq), all (fragmented) RNA molecules from a biological sample are sequenced. These sequences are then matched to the annotated genome sequence to identify to which gene each sequenced fragment belongs. More sequenced fragments mapping to a gene in one samples as compared to another, means that the specific gene had a higher activity. In addition, single cell sequencing technologies have allowed us to identify the heterogeneous nature of phenotypically pure populations and identity new sub populations of cells based on difference in gene expression.
In this workflow developed by Satija lab single cell transcriptomic data is used in combination of various dimensional reduction algorithms (such as PCA and t-SNE) to cluster cells. The leading genes in these clusterings can be used as genetic markers.
There is 1 package used in this workflow, which depend on 60 additional packages from CRAN (dependencies)
Used packages:
Package dependencies:
Single cell transcriptome data from Pollen et al. 2014