Sequence motif discovery with computational genome-wide analysis

As a result of the human genome project and advancements in DNA sequencing technology, we can utilize a huge amount of nucleotide sequence data and can search DNA sequence motifs in whole human genome. However, searching motifs with the naked eye is an enormous task and searching throughout the whole genome is absolutely impossible. Therefore, we have developed a computational genome-wide analyzing system for detecting DNA sequence motifs with biological significance. We used a multi-parallel network computing system as a powerful computing engine. Furthermore, we improved the system to work as a background engine for web-based applications. The multi-parallel computing engine consists of a head processor, which issues control commands to data processing nodes for various kinds of jobs, such as retrieving arbitrary sequences, generating mapping images, and loading data sections from genome databases. We constructed the system to function as a flexible Client/Server structure connected over the network, and this system could be adapted to cope with increases in sequence data and to deal with algorithms for new investigation needs by slightly changing the control procedures and increasing the number of the processor node. We developed two additional tools to annotate the genome sequences. The first was the cDNA Reverse Splicing Tool, which divided cDNA sequences into exons and mapped them on the genomic sequence, and the second was DNA-Protein Translation Tool which showed open reading frames (ORFs) of whole genome. In order to examine the availability and efficiency of our system, we searched and identified p53RE (p53 response element) as a representative sequence motif on genomic sequences of chromosome 21 and 22. As a result, we detected 50,000 p53REs on fifty mega base genomic DNA sequences within 27 seconds