On the visualization of the DNA sequence and its nucleotide content

Abstract

Visual inspection can help reveal patterns that would be computationally rather difficult to reveal. We consider three different algorithms for visualizations of a DNA sequence and its nucleotide content: random walk, fractal and visualization based on the entropy-like parameters calculated using a sliding window. We present a program that uses these three methods and visualizes either the whole of a given sequence, or specified fragments. It also provides facilities to compare visualizations obtained for different sequences/fragments. Random walk visualization considers the sequence symbol-by-symbol; the other two methods also take into account how well nucleotides are "mixed" in the sequence. It allows an easy visualization of repeated patterns, segments with a high/low content of some nucleotides, such as CG-islands, etc. The program also helps to identify regions of interest for further study.

Probabilistic NeuroScale for Uncertainty Visualisation

“…For human visual perception, the projection from the high dimensional space which holds all the data entities to the reduced dimensional form facilitates the understanding of the data and assists the interpretation by using the brain's ability to assimilate patterns in data. Clearly we can not perceive visualised images in more than three dimensions and can understand better two dimensional representations [5]. From the complicated structure of the data, trustworthy data transformation techniques are necessary to reduce data dimensionality to be used easily with any graphic representations for the end users.…”