Kalman Filtering Approach for Human Gene Finding

摘要

With the enormous amount of biological sequences that are available in the public domain, signal processing methods are playing an important role in genomic data processing. At first, we review the popular Voss representation which maps the alphabetic DNA sequence into the digital series. Secondly, in terms of the period-3 behavior of protein coding segments, the Kalman filtering approach for genomic signal processing is proposed, which can analyze and predict genes from real DNA sequences Thirdly, the presented approach is applied to identify the exons in a real human nucleotide sequence. It is shown that the developed method is useful for human gene identification by experiments.