Genomics & Bioinformatics

Biography

Biography: Jing Doo Wang

Abstract

With the progress of next generation sequences (NGS) nowadays, it is possible to have whole (complete) genomic sequences of instances of distinct organisms available. It is interesting and attractive to partition selected instances of organisms into classes according to features (phenotypes) defined or observed by domain experts precisely and then to extract and identify some distinctive genomic subsequences as biomarkers by comparing their whole genomic sequences. To overcome the computational bottleneck of whole genomic sequences comparison across those instances of organisms, a scalable approach based on previous work is applied to extract the maximal repeats from these tagged whole genomic sequences and meanwhile compute class frequency distributions of these maximal repeats extracted. These repeats with extremely biased class frequency distribution or just appearing in all instance of one class, if existing, may provide valuable hints or clues for biologists to further analyze or inspect whether the relationship between defined features (phenotypes) and extracted repeats (genotypes) is significant or not. Most of all, above computation of maximal repeat extraction could be achieved via cloud computing that can provide scalable computing environment if necessary. The method as described above opens a novel direction of researches to explore the connections between phenotypes and genotypes by comparing tagged whole genomic sequences.