Mesoscale transcriptome analysis

Abstract

In the current chapter, we will discuss some publications that applied the previously described data analysis methods to genome-scale expression datasets in order to investigate aspects of brain organisation at the mesoscale level (i.e. at the level of areas and their connections). The high-throughput technique for quantification of gene expression used in most of these works is the cDNA microarray that, until very recently, represented the technique of choice to obtain genome-scale gene expression data. The output of a microarray experiment is an n×r matrix, where n is the number of (oligonucleotide) probes printed in the microarray chip (note that multiple probes may be associated with a single mRNA) and r is the number of samples (brain regions) analyzed, and the elements of the matrix are normalised hybridisation signal intensities for the different probes (again, a single transcript may be represented by multiple probes). The structure of the dataset is very similar to the output of RNA-seq, so the down-stream analysis uses the same methods. A notable difference is that normalised signal intensities for microarrays are near-normally distributed.