Thanks, somehow with my software engineering mindset, I thought they somehow just copy&pasted the whole ARN sequence for encoding the spike protein from one virus to the other, and the end result should have been the same, but if its only parts of the S protein, then it makes sense that the outcome will be different.
If you want a software analogy, then I think it's more like using regular expressions to try to match inconsistent data, where new data (which is potentially inconsistent in new and exciting ways) is coming in all the time.
That’s a pretty good analogy the signature isn’t often not a simple hash of the malware but rather a more fuzzy signature based on core functions that are harder to alter without loss of function.
I don’t know if they even have enough room to encode the entire protein, that said even if they do it’s quite likely better to focus on specific amino acid chains that relate to specific structural areas of the protein that serve a critical function like interfacing with the receptors these aren’t likely to mutate without loss of function and it increases the likelihood of more targeted immune response.
If for example each vaccine selected 30 amino acid chains say 15 common ones and 15 that are varied between vaccines you can end up with vaccines that are more effective against certain mutations than others.
Realistically COVID-19 isn’t going anywhere and it will likely become very much endemic and will require a yearly vaccine especially for risk groups just like influenza.
