Replies

But look at it this way. It took 7 mutations for the protein to interact with cortisol. Calculate the odds of that randomly happening. However, those mutations had to occur in a certain order. Now calculate the odds.

I know I'm probably wasting my time but what the heck. Your post is a common example of fallacious probabilistic reasoning because the specification isn't independent. I'll try and explain by analogy although I'll simplify it somewhat.

You have 46 chromosomes, 23 pairs. You got one of each pair from your father and the other from your mother but the contributions are thought to be random. The probability you'd get exactly the set of chromosomes you did from your father is therefore 2^-23 and likewise from your mother. The probability of your genome is the square of that or 2^-46 or about 10^-14 which a little less likely than drawing a royal flush two times in a row. That's pretty damn small isn't it?

The probability of your grandparents producing exactly your genome is the square again or 10^-28. Got back two more generations and you're at 10^-112. Now creationist are fond of saying that there're only about 10⁸⁰ protons in the universe and "therefore" anything with a probability of less than 10^-80 simply can't happen. You're less probable and yet here you are. Right? Well, let's take that as a given even though you can't possibly actually exist according to the creationists.

The problem is the outcome you're calculating the probability of wasn't specified independently of your knowledge. The correct way to reason about the situation is to ask what is the probability of your great great grandparents having a great great grandchild, not you specifically (because that specification isn't independent) but some great great grandchild. And that's naturally quite a bit higher.

And similarly the right way to think about this mutation sequence is in terms of the universe of potential proteins you might get and how many of them are "interesting" which must be defined independently of your knowledge of the result. Achieving that kind of Independence is hard which is why this kind of "post" probability calculation is generally avoided.

But you are strictly talking about replication. Though there are many random outcomes, you have a basic set of rules and mechanisms for replication and the randomness works within those parameters.

It doesn’t deal with gene expression. In otherwards the machine is already built and we know it works, and we are only looking at the probabilities involved with one aspect of it’s functioning i.e. replication.

In evolution you are building the machine and creating the parameters for how it works, and doing it by random mutation. Not only do you have the probabilities involved in replication (in this case getting the right mutations to randomly happen and in the right order) but you have the probabilities of expression, i.e. that all those mutations will express themselves as a working organism.

Maybe I’m looking at it wrong but I see far, far more variables at work in evolution than simple gene replication of two viable genomes, and a far more complex set of probabilities.