ABSTRACT. In this paper, we run for all INDIA mutations and variants a biomathematical numerical method for analysing mRNA nucleotides sequences based on UA/CG Fibonacci numbers proportions (Perez, 2021). In this study, we limit ourselves to the analysis of whole genomes, all coming from the mutations and variants of SARS-CoV2 sequenced in India in 2020 and 2021. We then demonstrate - both on actual genomes of patients and on variants combining the most frequent mutations to the SARS-CoV2 Wuhan genomes and then to the B.1.617 variant - that the numerical Fibonacci AU / CG metastructures increase considerably in all cases analyzed in ratios of up to 8 times. We can affirm that this property contributes to a greater stability and lifespan of messenger RNAs, therefore, possibly also to a greater INFECTUOSITY of these variant genomes. Out of a total of 108 genomes analyzed: - None ("NONE") of them contained a number of metastructures LOWER than those of the reference SARS-CoV2 Wuhan genome. - Eleven (11) among them contained the same number of metastructures as the reference genome. - 97 of them contained a GREATER number of metastructures than the reference genome, ie 89.81% of cases. The average increase in the number of metastructures for the 97 cases studied is 4.35 times the number of SARS-CoV2 UA/CG 17711 Fibonacci metastructures. Finally, we put a focus on B.1.617.2 crucial exponential growth Indian variant. Then, we demonstrate, by analyzing the main worldwide 19 variants, both at the level of spikes and of whole genomes, how and why these UA / CG metastuctures increase overall in the variants compared to the 2 reference strains SARS-CoV2 Wuhan and D614G.