Data-Fusion for Epidemiological Analysis of Covid-19 Variants in UAE

Since December 2019, a new pandemic has appeared causing a considerable negative global impact. The SARS-CoV-2 first emerged from China and transformed to a global pandemic within a short time. The virus was further observed to be spreading rapidly and mutating at a fast pace, with over 5,775 distinct variations of the virus observed globally (at the time of submitting this paper). Extensive research has been ongoing worldwide in order to get a better understanding of its behaviour, influence and more importantly, ways for reducing its impact. Data analytics has been playing a pivotal role in this research to obtain valuable insights into understanding and fighting against the spread of infection. However, this is time and resource intensive, making it difficult to observe and quickly identify the impact of mutations. Factors such as the spread or virulence could explain the three months delay in revealing the new virus variant in the UK. This paper presents an extensive correlation analysis of the effect caused by the different SARS-CoV-2 strains, and their influence on the population across diverse factors, such as propagation and fatality rates, during the peak of the pandemic, with a focus on two major countries in the Middle East, the United Arab Emirate (UAE) and the Kingdom of Saudi Arabia (KSA). This research aims to investigate the epidemiological behaviour of the Coronavirus’ genomic variants over time in the UAE, compared with the KSA, where correlation analysis is carried out for a number of cases, deaths and their statistical deviations. The results of the analysis highlight very interesting insights into the epidemiological impact of the Covid-19 genomic behaviour in both countries, which could lead to important actions taken to minimize the impact on wider public health, possibly saving lives, and the economy. For instance, our method identifies a potential correlation between a spike in the number of deaths per case of 5.5 observed in the UAE by March 24th, with the emergence of new genomic variants of the Coronavirus (G0_c, G0_e1 and G0_e2). Our proposed methodology can be instrumental in identifying and classifying new variations of the virus earlier, and possibly predicting foreseeable mutations through pattern analysis, hence creating proactive measures to control its spread, such as the recent case of the new virus variant, recently discovered in the UK.