Chapter 14 - Predictive analytics of genetic variation in the COVID-19 genome sequence: a data science perspective

Abstract

Data obtained from genome sequencing have also contributed to several critical conclusions about the pathogen. Genomics, which deals with the organism's genetic material, is one of the scientific fields that is most promising for combating COVID-19. By unlocking the virus's genetic code and that of the hosts most seriously affected, experts intend to help advise public health decisions and discover successful care measures. The COVID-19 prevalence and death rates have been growing drastically, indicating a cumulative threat with increased age and co-occurring illnesses such as diabetes and cancer. Human gene characteristics can play a significant role in SARS—especially coronavirus—elevated transmissibility and epidemic severity. Data science enables the production of large-scale data sets of valuable observations. The corpus is added to sequence prediction models to estimate whether nucleotide basis (bases) are predictable from previous ones. An algorithm is used to identify positions in genome sequences where the nucleotide basis is altered and to determine the degree of mutation for mutation analysis of genome sequences. The primary objective of this analysis is to analyze the occurrence of genetic variations linked with primary immunodeficiencies in COVID-19 cases. To assess the applicability of gradient boosting tree (GBT) and neural network (NN), gradient neurons, XGBoost, convolution neural network, and pattern recognition combined with the entire genomic sequence study are used to provide fast and precise COVID-19 detections. XGBoost designs are used continuously to predict the applicability of GBT. Genetic factors occur over time and may contribute to the development of new varieties that may vary, improving the estimation of genome sequence codon families and codon pairs. The suggested approach is essential for detecting and mapping pathogenic virus outbreaks globally and local genetic variations.