Visual and Quantitative Analyses of Virus Genomic Sequences using a Metric-based Algorithm

Abstract

The worldwide spread of SARS-CoV-2 virus increases interest in the research of virus genomics and the creation of more advanced study methods. This work aims to develop a new fast DNA walk algorithm for one-dimensional visualization of RNAs based on a big-data method and comparative examination of several viruses and their lines and strains. In this work, a new metric-based algorithm for quantitative and visual analyses of RNAs is proposed and considered. It allows finding any fragments of genomic sequences using the Hamming distance between the binary-expressed RNA characters and symbols of a fragment under the search and building one-dimensional trajectories of genomic walks convenient for quantitative and qualitative analyses of RNAs and DNAs. Similarly, human-language texts can be processed and compared with genomic sequences.This algorithm is used to investigate the complete genomic sequences of SARS CoV-2, MERS, Dengue, and Ebola viruses available from Genbank and GISAID databases. The distributions of atg codon-starting triplets along with these sequences are built and considered as their atg-schemes. Additionally to the atg-walks, single-symbols distributions are calculated to detect the codon-content mutations, which do not change the atg-triplet coordinates along with genomic sequences. The visual analyses of distributions consisting of several hundred triplets enable us to define the level of stability of RNAs towards essential mutations and perform their classing. Statistical studies are applied to distributions of the inter-atg and inter-symbol distances along with genomic sequences. The fractal dimension values of these distributions are calculated, enabling them to correspond to the mutations discovered by Hamming walks and fractal-dimension values of several ten virus samples investigated here. The developed metric-based-based algorithm allows building one-dimensional RNA schemes of different scale levels and effectively analyzing the virus mutations with their classing.