PRFdb: A database of computationally predicted eukaryotic programmed-1 ribosomal frameshift signals

Background: The Programmed Ribosomal Frameshift Database (PRFdb) provides an interface to help researchers identify potential programmed -1 ribosomal frameshift (-1 PRF) signals in eukaryotic genes or sequences of interest. Results: To identify putative -1 PRF signals, sequences are first imported from whole genomes or datasets, e. g. the yeast genome project and mammalian gene collection. They are then filtered through multiple algorithms to identify potential -1 PRF signals as defined by a heptameric slippery site followed by an mRNA pseudoknot. The significance of each candidate -1 PRF signal is evaluated by comparing the predicted thermodynamic stability (Delta G degrees) of the native mRNA sequence against a distribution of Delta G degrees values of a pool of randomized sequences derived from the original. The data have been compiled in a user-friendly, easily searchable relational database. Conclusion: The PRFdB enables members of the research community to determine whether genes that they are investigating contain potential - 1 PRF signals, and can be used as a metasource of information for cross referencing with other databases. It is available on the web at http://dinmanlab.umd.edu/prfdb.