Transcription factor binding sites identification on human genome using an artificial neural network

Transcription factor binding sites on human DNA are the target locations of specific proteins called transcription factors. Gene expression process begins when a transcription factor binds to its target location in the genome. Expensive experimental methods are used to identify a limited number of these binding sites, hence there is essential need for computational algorithms. In this paper, we train a back propagation neural network to identify SP1 factor binding sites on human chromosome1. Biological data have been extracted from NCBI database which includes a wide variety of genetic information of human and other species. In order to compare the performance of our trained neural network with other classification algorithms, we use Support Vector Machine, Discriminant Analysis and K-Nearest Neighbor algorithm to classify same data. Results show that our trained neural network outperforms other classification algorithms.