Background

Much of the information in genome browsers regarding the structure of individual genes in the genomes of different organisms has been developed through strategies involving the implementation of exon-calling algorithms, coupled with mapping by homology with genes from other species. In general, this information has not been validated with experimental data, with the result often being that it is incomplete, inaccurate, or incorrect (see Figure 3). This has led to a quality-control problem for scientists who want to use this gene-specific information in their research. To address this issue, I have developed a simple experimental strategy that takes advantage of the large amounts of untapped gene expression data that have been deposited in the Sequence Read Archive of the National Center for Biotechnology Information (SRA NCBI), a searchable public resource that as of November 20, 2018 contains 8,548,792,923,294,171 nucleotides of open-access information from many different species of animals and plants. These data have been obtained from investigators who have used a variety of ‘next-generation’ DNA sequencing platforms to individually generate computer files containing tens of millions of base pairs of ‘RNA-sequencing’ results from a wide range of organisms, organs and tissues, developmental stages, and experimental paradigms. Here, I have taken advantage of the easy access to these data to describe a computational-based approach for mapping the 5’ and 3’ ends of genes, and for quantifying alternative RNA splicing. This series of experimental strategies offers effective replacements for older molecular biological methods, including combinations of cDNA cloning and PCR-generated approaches such as 5’ and 3’ RACE [rapid amplification of cDNA ends (Frohman et al., 1988)], and other more traditional assays [e.g., S1-nuclease and ribonuclease-protection mapping (Zinn et al., 1983)], and can be performed rapidly and reproducibly to help resolve the gene mapping problems noted above.