Neural Networks for Prediction of Nucleotide Sequences by using Genomic Signals

The conversion of symbolic sequences into complex genomic signals allows using signal processing methods for the handling and analysis of nucleotide sequences. This methodology reveals surprizing regularities, both locally and at a global scale, allowing us to predict nucleotides in a sequence, when knowing the preceding ones. Such experiments have a major biologic significance, as they explore the possibility and the efficiency of error correction in processes like replication, transcription and translation. Key-Words: Genomic signals, Nucleotide Sequences, Time series prediction, Sequence prediction, Neural networks