Re:Japanese scientists develop vaccine to eliminate cells be...
by pas on Sunday December 12, 2021 @03:07PM
(#29533102)
Do we know how exactly this happens? How fast this process "converges"? How big is the search space and how fast new candidates are generated?
Re:Japanese scientists develop vaccine to eliminate cells be...
by tux3 on Sunday December 12, 2021 @03:25PM
(#29533253)
About 3e11 combinations are possible, about 10 billions different antibodies will be generated by a human body (numbers I just sourced from Wikipedia).
The shuffling is done by recombination, which means joining segments of DNA in a random fashion (a similar recombination process is actually used for DNA repair, though with somewhat less randomness!)
There's a lot more to say about all of this, but I don't know how much immunology I can usefully describe in a few HN comments =]
Re:Japanese scientists develop vaccine to eliminate cells be...
by cipheredStones on Sunday December 12, 2021 @04:20PM
(#29533733)
(10 billion (1e9) / 3e11 ~= 3%, for those like me who don't have powers of 10 memorized.)
Does that imply that there's a good chance someone's immune system will just fail to generate any appropriate antibodies against a novel intruder? Or does the "fuzzy matching" aspect mean that exploring a small percentage of the space mean you're very likely to get at least some degree of success?
Someone else has mentioned the Kurzgesagt resources, which are great. If you want something more formal, OME is also a great resource I think.
https://onlinemeded.org/spa/immunology
>I always assumed the body generated matching snippets at random and the law of large numbers would result in a match eventually.
Right. Antibodies have a Y shape, where the two tips of the Y are variable, and the base is fixed. The IgM are the completely random snippets that happen to match a wide variety of things. However, because it's chemistry, it's more subtle than match or no match. Think of it like a variable degree of stickiness.
When a pathogen is found, your body is going to iteratively mutate new antibodies and select those that bind the best to the collected antigens, until you have something very specific (= sticks very readily). I'm simplifying a little bit, but that's the process for generating IgGs that I was talking about.