Getting poster data...
Srinivas Kudavelly (Philips Research Asia - Bangalore, Bangalore, India)Visualization of huge genomic data generated by next-generation sequencers represents big computational challenge. Spectral analysis by transforming the four signals (A, T, C, G) into frequency domain hold huge potential for visualizing this data. DNA spectrogram is generated by converting a DNA sequence to binary indicator sequence and then applying short-time Fourier transform (STFT) and mapping to a color space in order to visualize the output. We propose the visualization of huge genomic data in color space by using the dominant color in a frame (say 1024 bp). Magnitudes of a frequency component reveal certain patterns of nucleotides such as a larger value will indicates a stronger presence of repetitive sequence or di/ tri nucleotide such as CpG island, etc. The dominant colors of the successive genomic frame data can present slight differences. Output of the successive genomic frame data is smoothed until 'colorslider ’ is divided into well distinct color regions. The representation of a DNA sequence as a color bar allows patterns to be easily identified by visual inspection and automatic visual content analysis. These patterns will always include simpletandem repeats