Genomic sequencing generated an increased interest in non-coding DNA. This applies also to non-canonical DNA structures and their possible biological functions. Here we propose to combine molecular biology, bioinformatics and the latest computer technology in an effort to understand, predict and map the occurrence of biologically important DNA structures in genomes. Motivated by our previous research, we will focus on
non-canonical DNA structures (cruciforms, triplex, slippage- and Z-DNA). We will improve (and where necessary develop) tools for in silico prediction of these DNA structures. Predictions will be verified by carefully planned laboratory experiments, focusing on promoter regions of cancer-related genes (e.g. p53, MDM2, hsp90, EGFR) studied in the applicant's laboratory. Our ability to rapidly analyze full genomes will come from the use of special algorithms and applications built around FPGA hardware-acceleration cards. As a result, we will obtain annotations of genomes for predicted structures, occurence of such structures in oncogenes, as well as new methods and hardware for wider applications of accelerated sequence search.

Results

Publications