Recently advances in technology have made sequencing DNA easier, cheaper, and faster than ever. This leads to large pools of data on the actual sequence of the nucleotides present, but the larger problem presented by a sequence of DNA is understanding what it actually means, or whether it means anything at all. Gene prediction involves annotating a sequence of DNA to label introns, exons, and some softwares can even identify 5' untranslated regions. Most commonly, gene prediction is done on sequenced genomes. All of this is done with the objective of identifying the location of genes in the genome. Knowing which genes are present and where they are is vital in genetic research. Finding genes makes it easier to manipulate and study the genes, which can be applied in many types of research. Gene prediction is often useful in research as a first step to a complex project. This is because the gene prediction programs can pick out what segments of a genome are likely to contain genes, so there is less sequence left for more extensive testing. 

The Gene Prediction Waltz!

    An important characteristic of gene prediction programs is that each program is unique. This makes it possible to select a program that is the most appropriate for a particular project. Programs vary in the type of organism that they are optimized for, what computational programs are used to run them, and how 'trainable' they can be.