Message 374 of 393 (792901)
10-14-2016 10:39 PM
Reply to: Message 270 by Kleinman
10-14-2016 11:35 AM
Re: Mathematics cannot change reality and when done incorrectly predicts nothing
| ... What I do remember is that they listed at least 8 genes necessary to be transformed. ... |
Excuse me for coming in late in the discussion. I have received your pdfs but not had time to look them over yet. Can you answer a simple question for me, even if it has already been asked?
What is the probability that a mutation will be beneficial?
Your question hasn't been asked and my answer is I don't know. ...
If you don't know then you cannot invalidate evolution with your calculations.
| ... But this is not a number which you have to know to understand how rmns works.|
Evolution works by beneficial mutations being selected by natural processes -- success in living and breeding. The probability of a mutation being beneficial would seem to be central to any mathematical approach trying to show that evolution is broken.
I think we can all agree that mutations are random -- leaving aside for the moment that the probability of mutations varies with the section of DNA involved -- and that some are immediately deleterious or immediately beneficial, while others are immediately neutral and their relative deleterious\beneficial value can be important later.
We also have cases where a mutation is somewhat deleterious but leads later to beneficial results because of changing environmental conditions.
So how can we predict the probability of a mutation being beneficial?
Most people say that most mutations are neutral. ...
Which leaves the door open to later mutations that can build on them.
| ... Mutations are fairly rare to begin with. ... |
Yet every individual in every species has several.
| ... Most DNA replication is done with high fidelity. ... |
For individuals that survive from zygote to born young. All those that die from birth defects are because of less than sufficient fidelity.
| ... As you read my papers on rmns, you will see that I address the possibility that even though a mutation occurs at the correct site in a genome, it has to be the correct mutation to improve fitness. ... |
This is confused. There is one mutation, it occurs randomly, and that means both location and format\type are part of the same mutation, not two separate problems. The same type in a different location would be a different mutation.
There is no "correct mutation" -- the mutation happens and then selection operates on that mutation, whether it is in location A or location B, whether it is type K or type L. Whether or not it is beneficial is that probability discussed above that you admit you don't know.
This seems to be the root of your problem, trying to make a single mutation event into a two event process.
It also seems from this that you are calculating the probability of a given mutation occurring in a second individual. Certainly when you go to two mutations occurring independently in different individuals the maths would give an extremely low probability for occurrence, but that is not how evolution works.
| ... Just getting an accurate mutation rate is a challenging problem and then determining the fraction of the mutations which are beneficial, neutral and detrimental is even more challenging. ... |
Yet we know that they all occur every generation of every species. All natural selection needs are some beneficial mutations and a low rate of death/fertility problems (where selection pressure enters the picture).
| ... But the mutation rate is not the dominant factor in the rmns problem, it is the multiplication rule of probabilities that drives this phenomenon. ... |
This too is confused. If I take a coin and toss it 53 times I end up with a pattern of heads and tails, and the probability of my getting that specific pattern is 1.
If I try to match that pattern with another 53 tosses the probabilities are, by the multiplication rule, extremely minute. You only use multiplication when the same steps need to be reproduced. Evolution does not work that way.
| ... It is the joint probability that two or more beneficial mutation occur on a lineage which drives this problem.|
Again, we get back to the question of the probability that a mutation will be beneficial, which you admitted you don't know ... certainly then you can't know the probability of a second mutation being beneficial, but that isn't the worst of your problem.
There are actual documented experiments (one involving E. coli) where a neutral mutation occurs in one generation and then in a later generation a second mutation occurs where the combination is beneficial, meaning that the original mutation is now beneficial. Calculating the probability that those two specific mutations would occur (the "correct mutations" at the "correct locations") would result in a very small number, but the probability that it occurred is 1: it happened.
Your model is wrong because there is an assumption of structure to the mutation process being necessary to evolution, and that assumption is false.
|This message is a reply to:|
| ||Message 270 by Kleinman, posted 10-14-2016 11:35 AM|| ||Kleinman has not yet responded|