Promoter2.0: for the recognition of PolII promoter sequences

Motivation : A new approach to the prediction of eukaryotic PolII promoters from DNA sequence takesadvantage of a combination of elements similar to neural networks and genetic algorithms to recognize a set ofdiscrete subpatterns with variable separation as one pattern: a promoter. The neural networks use as input a smallwindow of DNA sequence, as well as the output of other neural networks.

Through the use of geneticalgorithms, the weights in the neural networks are optimized to discriminate maximally between promoters andnon-promoters. Results : After several thousand generations of optimization, the algorithm was able todiscriminate between vertebrate promoter and non-promoter sequences in a test set with a correlationcoefficient of 0.63.

In addition, all five known transcription start sites on the plus strand of the completeadenovirus genome were within 161 bp of 35 predicted transcription start sites. On standardized test setsconsisting of human genomic DNA, the performance of Promoter2.0 compares well with other softwaredeveloped for the same purpose.

Availability : Promoter2.0 is available as a Web server at http://www.cbs.dtu.dk/services/promoter/ Contact : steen@cbs.dtu.dk