/sci/ - Science & Math

>>8807867
You are correct of course. There are probabilistic aspects to this. We would definitely not expect mutations to produce a person that is born from a watermelon even if it's theoretically possible.

This is just how the conversation is directed from "it can't happen" to discussing the probabilities of mutations and how the steps of evolution take place. It's important to understand that evolution works by changing the letters of the DNA (mostly perhaps one at a time) and not by magically transforming a fish into a frog.

The next step is to look at the number of mutations per generation which in humans is around 130. That means it would take 25 million generations to get as many mutations as there are nucleotides in the genome. OR 25 million people in 1 generation. Assuming humans have had a population size of 1 million for the past 1 million years and a generation time of 25 years (40 000 generations in 1 M years) we get around 5200 billion mutations. In a genome of 3,2 billion letters, that's a lot.

If we look at chimps and humans who differ by 1 %, or around 32 million letters, and find that our species separated 5 million years ago, and assume 1 million population size for both lineages and a generation time of 20 years, we get around 65 000 billion mutations. So maybe 1 out of every 2 million mutations got stuck in our genomes.

To me it looks like there's lots of opportunities for mutations to happen and to spread.