Eukaryotic Gene Prediction

Introduction: The advent of large-scale genome sequencing has revolutionized the field of genetics and biology. Sequencing projects require sophisticated computational analysis to manage vast collections of data. Scientists first sequenced a genome in 1977, that of a small bacteriophage consisting of 11 genes over 5.4kb of DNA. In the bacteriophage, coding genes comprise 95% of the genome. 1 Since then, numerous prokaryotic and eukaryotic genomes have been sequenced, including the mouse (M. musculus) genome, the human (H .sapiens) genome and the model plant Arabidopsis thaliana genome. Gene prediction in eukaryotic genomes can be especially difficult given the large genome size, the low proportion of coding regions, and the frequent splice events due to the presence of introns (non-coding segments between the exons that code for a gene). For example, the human genome contains approximately 25,000 genes over 30 million base pairs. Given that the average protein encoded by these genes is 350 amino acids, this means that only about 1% of the genome actually codes for proteins and these regions must be separated from the remaining 99% of the genome. 1 The actual coding regions are frequently interrupted by introns that are removed from the mRNA transcripts through splicing.