Drop-seq at Warwick

Thanks to a donation we have recently purchased a device for single-cell transcriptome barcoding. Using it one can attach cell-specific molecular barcodes to mRNA sequences in about 5 percent of cells out of a sample of about 100k cells. These can then be processed and sequenced. The data coming out of this is essentially a matrix with read numbers for all genes in all successfully barcoded cells. These transcriptomes are typically very sparsely represented by few reads, but using techniques such as t-SNE (t-distributed stochastic neighbour embedding) related cells and cell types can readily be identified, cells can be grouped and data from many cells of the same type can be pooled. I am going to explain the device, present sample data sets, and discuss the statistical analysis.