Is the Random Mutation Problem in Evolution a Valid Criticism?

[big_bounce]

"He provided empirical backing for this conclusion from experimental research he earlier published in the Journal of Molecular Biology, finding that only one in 10⁷⁴ amino-acid sequences yields functional protein folds."

1.Above sentence means in 10⁷⁴ mutations(Events) just one of them could make a functional protein and that mutation is beneficial?

2.Is above sentence similar to "Hoyle's calculation of the Odds of Life", thus the criticism against Hoyle's idea can apply to above sentence?
Like this:

That said, there are several problems with this analysis:

• Further, Hoyle's calculations, as described by many Christians, assume that the combinations are independent events, which is not generally the case with molecular interactions. Lastly, it is not valid to apply probability to an event that has already occurred -- it's like saying that the odds of winning the lottery are 1 in 23 million, so no one could have won it.
• The analysis assumes independence of variables (proteins, molecules, etc.) This is obviously false, as we know that proteins strongly interact. This means that certain combinations will not occur because the interactions disallow them. If one takes dependence into account, the numbers will be lower.
• The analysis assumes independence of trials. That is, it assumes that one combination is created, evaluated, then destroyed. In fact, the combinations interact strongly with each other---groups of proteins interact with other groups of proteins. As a result, the analysis is flawed.
• Even if we believe the independence argument, we have no data on how many trials there have been. In the lottery, we know that someone will win because we know that even though the odds are astronomical of winning, there are many tickets sold. When it comes to life, we have no idea how many "losing tickets" there are. So any number is meaningless unless we have the context for interpreting it.
• http://www.coppit.org/god/hoyle.php

 

[Simon Bridge]

1. I'm reading it as a statement of probability. A 1/e74 probability that a random variation in whateveritwas produces a "functional protein fold".
That means that if you have a large number of whateveritwas's M, and you count the number of functionally folded proteins N, then N/M will approach 1/e74 as M gets very large.

If the probability of a success is p=1/n then it is tempting to think we need n trials to get one success ... but that is not the case. You can get a success in just one trial. To see what's going on you can work out the probability of at least one success if there are n trials ... or work out the number of trials m you need so the probability of at least one success is some value you feel good about (0.5, 0.99...).

2. Need more data... when you quote something, please provide a citation.
Whether the criticisms apply depends on how the probability was arrived at.

We can guess: it is claimed to be from empirical data ... so if the probability is measured, then it does not suffer from the flaws in Hoyles argument. But some uses of the figure may do.

 

[Drakkith]

1.Above sentence means in 10⁷⁴ mutations(Events) just one of them could make a functional protein and that mutation is beneficial?

Do you have a reference for the first quote?

 

[Ygggdrasil]

Above sentence means in 1074 mutations(Events) just one of them could make a functional protein and that mutation is beneficial?

No. If you write a random sequence of amino acids, very few will fold into stable structures. However, extant protein sequences are not random strings of amino acids, they were selected by evolution to lie in an area of sequence space that form stable folds. Therefore, many mutations to extant proteins have no effect (I think the number is ~50%, but I'll have to check my references later)

 

[big_bounce]

Do you have a reference for the first quote?

http://www.evolutionnews.org/2014/10/a_reader_asks_c090811.html

In a 2010 peer-reviewed study, molecular biologist Douglas Axe demonstrated the inability of Darwinian evolution to produce multi-mutation features. Axe calculated that when a "multi-mutation feature" requires more than six mutations before giving any benefit, it is unlikely to arise even in the whole history of the Earth.4 He provided empirical backing for this conclusion from experimental research he earlier published in the Journal of Molecular Biology. He found there that only one in 1074 amino-acid sequences yields a functional protein fold.5 That implies that protein folds in general are multi-mutation features, requiring many amino acids to be fixed before the assembly provides a functional advantage

If the probability of a success is p=1/n then it is tempting to think we need n trials to get one success ... but that is not the case. You can get a success in just one trial.

I know it, But i said we need a genetic code for making a functional protein.
The probability of generate this code by random mutation is 1 in 10⁷⁴?

No. If you write a random sequence of amino acids, very few will fold into stable structures. However, extant protein sequences are not random strings of amino acids, they were selected by evolution to lie in an area of sequence space that form stable folds. Therefore, many mutations to extant proteins have no effect (I think the number is ~50%, but I'll have to check my references later)

1.All of proteins in organisms that selected by evolution are a functional protein.
2.Probability of find a functional protein is 1 in 10⁷⁴.
3.DNA codes proteins and make them.
Then probability a random mutation for codes a functional protein is 1 in 10⁷⁴.

 

[Drakkith]

Your references are from an intelligent design website, and Douglas Axe himself, the author of the paper claiming the 1 in 10⁷⁴ chance, is a proponent of intelligent design. While that doesn't automatically make his assertions incorrect, it does cast serious doubt on their validity.

This puts this thread on thin ice. I'm willing to leave it open provided the discussion remains on valid science, but any deviation from that will result in this thread being locked.

 

[Ygggdrasil]

1.All of proteins in organisms that selected by evolution are a functional protein.
2.Probability of find a functional protein is 1 in 1074.
3.DNA codes proteins and make them.
Then probability a random mutation for codes a functional protein is 1 in 1074.

A random mutation is different from a random sequence. As an analogy, consider that about 70% of the Earth's surface is covered in water. If you drop a person on a random spot on the globe, the chance they are landing in water is 70%. However, this does not mean that every time you take a step, you have a 70% chance of landing in water. Areas covered by water cluster together on the surface of the Earth as do areas covered by land. If you are on an area covered by land, you have a high probability of finding another area covered by land for each step you take.

Similarly, if you are starting off at a functional protein fold, then mutating that protein (i.e. changing the identity of one amino acid) is likely to keep you in the area of functional protein folds (according to experimental data, ~62% of amino acid mutations have no effect on function, ~30% are deleterious to function, and 8% appear to be beneficial to function. See http://www.nature.com/nrg/journal/v11/n8/full/nrg2808.html). So once functional protein folds have been found, evolution has ways of diversifying these folds to find other functional protein structures.

Of course, this begs the question of how the first set of functional protein folds came about in the first place. This is related to work on abiogenesis (the origin of life) and the question is not well understood. It's possible that life began with RNA as the functional molecule, and the first protein-like structures were small peptides that acted as scaffolds to bind and help the RNA molecules to retain their structure. Over time, these small peptides could come to take up more of the structure until eventually you could replace the RNA entirely with protein. It's also worth noting that there are some intrinsically unstructured proteins that still have important biological functions despite not folding into any defined structure. Furthermore, various "low complexity" sequences also have functional roles that rely less on having the protein fold into a specific 3D shape, and could provide a potential route for functional peptides to begin to coalesce into protein-like structures.

 

[Laroxe]

I'm aware there is some current work going on that suggests that mutations might not be quite as random as was thought, one such idea is a model called Nutrient-dependent/pheromone-controlled adaptive evolution: and there is a lot of interest in the role of viruses in inducing specific changes. As yet I haven't got my head around either but there is quite a lot of literature around about both.