The emergence of a heavily mutated SARS-CoV-2 variant (B.1.1.529, Omicron) and it’s spread to 6 continents within a week of initial discovery has set off a global public health alarm. Characterizing the mutational profile of Omicron is necessary to interpret its shared or distinctive clinical phenotypes with other SARS-CoV-2 variants. We compared the mutations of Omicron with prior variants of concern (Alpha, Beta, Gamma, Delta), variants of interest (Lambda, Mu, Eta, Iota and Kappa), and all 1523 SARS-CoV-2 lineages constituting 5.4 million SARS-CoV-2 genomes. Omicron’s Spike protein has 26 amino acid mutations (23 substitutions, two deletions and one insertion) that are distinct compared to other variants of concern. Whereas the substitution and deletion mutations have appeared in previous SARS-CoV-2 lineages, the insertion mutation (ins214EPE) has not been previously observed in any SARS-CoV-2 lineage other than Omicron. The nucleotide sequence encoding for ins214EPE could have been acquired by template switching involving the genomes of other viruses that infect the same host cells as SARS-CoV-2 or the human transcriptome of host cells infected with SARS-CoV-2. For instance, given recent clinical reports of co-infections in COVID-19 patients with seasonal coronaviruses (e.g. HCoV-229E), single cell RNA-sequencing data showing co-expression of the SARS-CoV-2 and HCoV-229E entry receptors (ACE2 and ANPEP) in respiratory and gastrointestinal cells, and HCoV genomes harboring sequences homologous to the nucleotide sequence that encodes ins214EPE, it is plausible that the Omicron insertion could have evolved in a co-infected individual. There is a need to understand the function of the Omicron insertion and whether human host cells are being exploited by SARS-CoV-2 as an ‘evolutionary sandbox’ for host-virus and inter-viral genomic interplay.