Probably the key point is here. > The probability is calculated based on the neutral theory of molecular evolution and binominal distribution of non-synonymous

Probably the key point is here. > The probability is calculated based on the neutral theory of molecular evolution and binominal distribution of non-synonymous (N) and synonymous (S) mutations. This means (ELI15): We use as a starting point the belief ("a theory" [0]) that natural mutations do not simply occur randomly, but instead follow certain rules. The fact that there are rules, makes certain combinations of mutations (i.e. not just single mutations but the case where mutations abc1234, xd456, er6523, and sdf374 happen together) more or less likely than other combinations of mutations. Given that theory, we calculate how likely the particular combination of mutations included in the Omicron variant is to occur naturally. It turns out it's quite unlikely. --- Not defending or attacking the paper or its conclusions, as I haven't read it. Just explaining the difference to randrange(1000000) which is a. just one single item; b. all items equally likely even if called repeatedly. --- [0] https://en.wikipedia.org/wiki/Neutral_theory_of_molecular_evolution https://en.wikipedia.org/wiki/Neutral_theory_of_molecular_ev... - "A neutral mutation is one that does not affect an organism's ability to survive and reproduce. The neutral theory assumes that most mutations that are not deleterious are neutral rather than beneficial."