Читать книгу Algorithms in Bioinformatics - Paul A. Gagniuc - Страница 25

The set of all RNA molecules produced by a given organism is known as the “transcriptome.” This includes, of course, the mRNA transcripts but also the RNA molecules mentioned above (i.e. mRNAs, tRNAs, rRNAs, siRNAs, miRNAs, piRNAs, and so on) as well as other uncharacterized noncoding RNA molecules. When expressed, genes produce mRNAs in different quantities, which are then detectable [69]. Currently, two main techniques are representative for capturing gene expression, namely: RNA-Seq and microarrays [70, 71]. RNA-Seq (RNA sequencing) allows for full sequencing of all RNA molecules present in a sample, whereas microarrays target known transcripts of different genes through hybridization (complementary) [72]. Thus, RNA-Seq experiments can estimate the subset of genes expressed in a cell type or in different tissues (several cell types) at any one time by an alignment of the sequenced RNAs to the reference genome (the DNA of the organism) [73]. However, the transcriptome can be seen as an ideal set, because the complete set of possible RNAs cannot be fully detected. Reasoning dictates that each state of a cell shows a specific subset of RNAs from the transcriptome. Of the total number of states that a cell can exhibit, only a few states can be induced and captured by RNA-Seq. Thus, a small subset of RNAs from the transcriptome may remain undetectable. At the tissue level, there are a number of cell types, each with a specific set of active genes. Often, the analysis of the pattern of gene expression is performed at the tissue level, i.e. on several cell types at the same time. From a global perspective, this leads to a union between the sets of genes expressed in each of the cell types that make up the tissue. Furthermore, genes that are expressed in several cell types (such as housekeeping genes) may show the highest amounts of mRNA, while genes that are only expressed in certain cell types can show lower amounts of mRNA.