DataBase of Genome Periodicity
A nonredundant catalog of DNA periodicity regions in a number of complete genomes of model eukaryotes presents the results of an automated spectral-statistical approach to genome-wide periodicity search. Significance level of the revealed heterogeneities is 10-6 and below. Heterogeneity regions listed in the database are potential regions of latent periodicity, indicated mainly by approximate tandem repeats. A specified Period Length, in which statistically significant heterogeneity has been determined, presents an estimation of periodicity pattern size. The Exponent shows the number of copies for this pattern in the revealed periodicity region. Average invariance for copies of the estimated periodicity pattern is characterized by a Preservation Level (ranging from 0.3 to 1.0) for the characters across period positions. The new approach was implemented to reveal periodicities, presented in the current version of Database.
A special record unit – known as a Group – has been developed for the database. A group contains associated overlapping heterogeneity regions, of which the longest one serves as group representer. Group members lying inside this representer are regarded as its intrinsic heterogeneities. These indicate fragments of the representer that are more structured and facilitate correct data interpretation. Some groups contain only one region: the group representer. Different groups do not intersect. In accordance with the two-level group structure (a representer and the intrinsic heterogeneities), there are two Output modes: nonredundant (only representers will be shown) and simple (all regions in the database that meet the specified search parameters will be shown).