Yale University computer science professor takes a look at protein formation probabilities

When I was in graduate school, we studied a book called “Mirror Worlds”, authored by famous computer science professor David Gelernter at Yale University. This week, I noticed that Dr. Gelernter had written an article in the prestigious Claremont Review of Books. In his article, he applies his knowledge of computer science to the problem of the origin of life.

Evolution, if it is going to work at all, has to explain the problem of how the basic building blocks of life – proteins – can emerge from non-living matter. It turns out that the problem of the origin of life is essentially a problem of information – of code. If the components of proteins are ordered properly, then the sequence folds up into a protein that has biological function. If the sequence is not good, then just like computer code, it won’t run.

Here’s Dr. Gelernter to explain:

How to make proteins is our first question. Proteins are chains: linear sequences of atom-groups, each bonded to the next. A protein molecule is based on a chain of amino acids; 150 elements is a “modest-sized” chain; the average is 250. Each link is chosen, ordinarily, from one of 20 amino acids. A chain of amino acids is a polypeptide—“peptide” being the type of chemical bond that joins one amino acid to the next. But this chain is only the starting point: chemical forces among the links make parts of the chain twist themselves into helices; others straighten out, and then, sometimes, jackknife repeatedly, like a carpenter’s rule, into flat sheets. Then the whole assemblage folds itself up like a complex sheet of origami paper. And the actual 3-D shape of the resulting molecule is (as I have said) important.

Imagine a 150-element protein as a chain of 150 beads, each bead chosen from 20 varieties. But: only certain chains will work. Only certain bead combinations will form themselves into stable, useful, well-shaped proteins.

So how hard is it to build a useful, well-shaped protein? Can you throw a bunch of amino acids together and assume that you will get something good? Or must you choose each element of the chain with painstaking care? It happens to be very hard to choose the right beads.

Gelernter decides to spot the Darwinist a random sequence of 150 elements. Now the task the Darwinist is to use random mutation to arrive at a sequence of 150 links that has biological function.

[W]hat are the chances that a random 150-link sequence will create such a protein? Nonsense sequences are essentially random. Mutations are random. Make random changes to a random sequence and you get another random sequence. So, close your eyes, make 150 random choices from your 20 bead boxes and string up your beads in the order in which you chose them. What are the odds that you will come up with a useful new protein?

[…]The total count of possible 150-link chains, where each link is chosen separately from 20 amino acids, is 20150. In other words, many. 20150 roughly equals 10195, and there are only 1080 atoms in the universe.

What proportion of these many polypeptides are useful proteins? Douglas Axe did a series of experiments to estimate how many 150-long chains are capable of stable folds—of reaching the final step in the protein-creation process (the folding) and of holding their shapes long enough to be useful. (Axe is a distinguished biologist with five-star breeding: he was a graduate student at Caltech, then joined the Centre for Protein Engineering at Cambridge. The biologists whose work Meyer discusses are mainly first-rate Establishment scientists.) He estimated that, of all 150-link amino acid sequences, 1 in 1074 will be capable of folding into a stable protein. To say that your chances are 1 in 1074 is no different, in practice, from saying that they are zero. It’s not surprising that your chances of hitting a stable protein that performs some useful function, and might therefore play a part in evolution, are even smaller. Axe puts them at 1 in 1077.

In other words: immense is so big, and tiny is so small, that neo-Darwinian evolution is—so far—a dead loss. Try to mutate your way from 150 links of gibberish to a working, useful protein and you are guaranteed to fail. Try it with ten mutations, a thousand, a million—you fail. The odds bury you. It can’t be done.

Keep in mind that you need many, many proteins in order to have even a simple living cell. (And that’s not even considering the problem of organizing the proteins into a system).

So, if you’re a naturalist, then your only resources to explain the origin of life are chance and mutation. As Dr. Gelernter shows, naturalistic explanations won’t work to solve even part of the problem. Not even with a long period of time.  Not even if you use the entire universe as one big primordial soup, and keep trying sequences for the history of the universe. It just isn’t possible to arrive at sequences that have biological function in the time available, using the resources available. The only viable explanation is that there is a computer scientist who wrote the code without using trial and error. Something that ordinary software engineers like myself and Dr. Gelernter do all the time. We know what kind of cause is adequate to explain functioning code.

Yale University computer science professor takes a look at protein formation probabilities

When I was in graduate school, we studied a book called “Mirror Worlds”, authored by famous computer science professor David Gelernter at Yale University. This week, I noticed that Dr. Gelernter had written an article in the prestigious Claremont Review of Books. In his article, he applies his knowledge of computer science to the problem of the origin of life.

Evolution, if it is going to work at all, has to explain the problem of how the basic building blocks of life – proteins – can emerge from non-living matter. It turns out that the problem of the origin of life is essentially a problem of information – of code. If the components of proteins are ordered properly, then the sequence folds up into a protein that has biological function. If the sequence is not good, then just like computer code, it won’t run.

Here’s Dr. Gelernter to explain:

How to make proteins is our first question. Proteins are chains: linear sequences of atom-groups, each bonded to the next. A protein molecule is based on a chain of amino acids; 150 elements is a “modest-sized” chain; the average is 250. Each link is chosen, ordinarily, from one of 20 amino acids. A chain of amino acids is a polypeptide—“peptide” being the type of chemical bond that joins one amino acid to the next. But this chain is only the starting point: chemical forces among the links make parts of the chain twist themselves into helices; others straighten out, and then, sometimes, jackknife repeatedly, like a carpenter’s rule, into flat sheets. Then the whole assemblage folds itself up like a complex sheet of origami paper. And the actual 3-D shape of the resulting molecule is (as I have said) important.

Imagine a 150-element protein as a chain of 150 beads, each bead chosen from 20 varieties. But: only certain chains will work. Only certain bead combinations will form themselves into stable, useful, well-shaped proteins.

So how hard is it to build a useful, well-shaped protein? Can you throw a bunch of amino acids together and assume that you will get something good? Or must you choose each element of the chain with painstaking care? It happens to be very hard to choose the right beads.

Gelernter decides to spot the Darwinist a random sequence of 150 elements. Now the task the Darwinist is to use random mutation to arrive at a sequence of 150 links that has biological function.

[W]hat are the chances that a random 150-link sequence will create such a protein? Nonsense sequences are essentially random. Mutations are random. Make random changes to a random sequence and you get another random sequence. So, close your eyes, make 150 random choices from your 20 bead boxes and string up your beads in the order in which you chose them. What are the odds that you will come up with a useful new protein?

[…]The total count of possible 150-link chains, where each link is chosen separately from 20 amino acids, is 20150. In other words, many. 20150 roughly equals 10195, and there are only 1080 atoms in the universe.

What proportion of these many polypeptides are useful proteins? Douglas Axe did a series of experiments to estimate how many 150-long chains are capable of stable folds—of reaching the final step in the protein-creation process (the folding) and of holding their shapes long enough to be useful. (Axe is a distinguished biologist with five-star breeding: he was a graduate student at Caltech, then joined the Centre for Protein Engineering at Cambridge. The biologists whose work Meyer discusses are mainly first-rate Establishment scientists.) He estimated that, of all 150-link amino acid sequences, 1 in 1074 will be capable of folding into a stable protein. To say that your chances are 1 in 1074 is no different, in practice, from saying that they are zero. It’s not surprising that your chances of hitting a stable protein that performs some useful function, and might therefore play a part in evolution, are even smaller. Axe puts them at 1 in 1077.

In other words: immense is so big, and tiny is so small, that neo-Darwinian evolution is—so far—a dead loss. Try to mutate your way from 150 links of gibberish to a working, useful protein and you are guaranteed to fail. Try it with ten mutations, a thousand, a million—you fail. The odds bury you. It can’t be done.

Keep in mind that you need many, many proteins in order to have even a simple living cell. (And that’s not even considering the problem of organizing the proteins into a system).

So, if you’re a naturalist, then your only resources to explain the origin of life are chance and mutation. As Dr. Gelernter shows, naturalistic explanations won’t work to solve even part of the problem. Not even with a long period of time.  Not even if you use the entire universe as one big primordial soup, and keep trying sequences for the history of the universe. It just isn’t possible to arrive at sequences that have biological function in the time available, using the resources available. The only viable explanation is that there is a computer scientist who wrote the code without using trial and error. Something that ordinary software engineers like myself and Dr. Gelernter do all the time. We know what kind of cause is adequate to explain functioning code.

Video, audio and summary of intelligent design discussion, with Ben Shapiro and Stephen C. Meyer

(Re-posting this for those who missed it the first time around)

This is a must-see / must-hear conversation. Famous conservative Ben Shapiro discusses evolution and intelligent design with Dr. Stephen C. Meyer, author of “Signature in the Cell” (origin of life) and “Darwin’s Doubt” (fossil record). These are the two best books on intelligent design. Their conversation covers the most important issues in the origins debate.

So to start, here’s the full video:

Someone actually put out an outline on Youtube, with a time-based index, so I’ll just steal that as a base, and add to it:

• 1:34 What is your scientific background? Science undergraduate degree, professional geologist, later did a PhD in philosophy of science from Cambridge University.
• 2:39 What is the difference between intelligent design and creationism? Creationism starts from the Bible and posits a shorter history of the universe. ID starts from the data of the natural world and is neutral about the age of the Earth / universe. Meyer accepts the old-Earth.
• 3:36 How is Intelligent Design a scientific theory? The discovery of DNA reveals that code is central to living systems. Intelligent design uses the method of “inference to the best explanation” in order to argue that the best explanation for the code is an intelligent agent.
• 6:10 What evidence would have to arise to make Intelligent Design Falsifiable? If a naturalistic mechanism was discovered that could produce biological information using the probabilistic resources of the universe, and the time available, then intelligent design would be falsified.
• 7:26 Is religion separate from science or intertwined within it? There are three views: science is totally separate from religion, science is in total conflict with religion, science and religion agree on some issues, e.g. – the origin of the universe and Genesis 1:1. There are areas where science and religion overlap.
• 9:55 Why are the most prominent Darwinians also militant atheists? Evolution is a theory that tries to explain nature using naturalistic mechanisms, so it is compatible with atheism.
• 10:45 What does the theory of evolution say? The term evolution has multiple meanings, and should be defined before discussions. It can refer to change over time. It can refer to animals changing slightly to adapt to enviromental changes. It can refer to the idea that all animals evolved from simpler life forms, and that there is a tree of life showing how different types of organisms share common ancestors. And it can refer to the idea that purely undirected processes can explain the history of life using purely materialistic forces. It’s that final view that intelligent design challenges.
• 13:15 Where is the discontinuity in naturalistic processes in the development of life? The first discontinuity is the origin of simple life from non-living components. The second discontinuity is the sudden appearance of different body plans in a very narrow window of time in the Cambrian era.
• 15:42 Why does information theory suggest that code requires some sort of designer? DNA is a true information-bearing system identical to the software in computers, e.g. – operating system, applications.
• 19:45 Can information be created by random mutation, and favorable mutations preserved by natural selection? Just as in software code, instructions must be added in order to develop new functionality. Random additions of characters will almost always degrade biological function. The number of possible sequences that do nothing useful is vastly higher than the number of sequences that perform biological functions. Doug Axe did research on this at Cambridge University, and he found that the number of functional sequences of amino acids is 1 in 10 to the 77th power. Given the probabilistic resources (replicating organisms)and the time available, it is extremely unlikely to find sequences that have functional information by chance.
• 25:05 What about Stephen Jay Gould’s model of punctuated equilibrium – doesn’t it explain the sudden jumps in information? Gould’s mechanism is accurate according to the fossil record, which shows a lot of jumps. But he did not have a naturalistic explanation for sudden jumps in biological function. Darwinian mechanisms work slowly and would (in theory) produce different body plans gradually. But this is not what the fossil record shows.
• 27:22 What is the mechanism for injecting information in the theory of intelligent design? The information comes in from an intelligence when new major body plans appear, and minor variations within types could be explained by evolution.
• 29:25 Does the Miller-Urey experiment provide a naturalistic explanation for the building blocks necessary for the origin of life? The MU experiment only produced a few types of amino acids, it doesn’t say anything about how to sequence the amino acids in order to form protein folds that can perform biological functions. The MU experiment also pre-supposes conditions on the early Earth (reducing gases) that do not match what was there (oxidyzing or neutral gases).
• 32:00 Is the RNA world hypothesis is a good explanation for the origin of life? Evolution requires that replication already be in place, because evolution assumes that mutations appear during the replication. The RNA world hypothesis suggests that sequences contain information, but also catalyze origin of life chemistry. The problem with RNA world is that it starts with self-replicating systems. And those replicating systems require the scientist to inject information into the system to get even the simplest replication started.
• 34:56 How do scientists respond to the critiques of Darwinism proposed by intelligent design advocates? By and large, they accept them. They think that mutation and selection works once living systems are in place, but they realize it has no explanation for the origin of life or the sudden origin of body plans. (Tells about the  conference of the Royal Society, where problems with Darwinian mechanisms were discussed, and the 2003 MIT Press book by Muller and Newman).
• 37:16 Why do people hold to Darwinian evolution in the face of these problems? Many scientists presuppose methodological naturalism, which requires that any explanation for the origin of life and the origin of major body plans involve materialist explanations only. No intelligent agents are allowed. The problems occur when assumption of naturalism causes scientists to propose incorrect explanations for what we observe in nature. It’s also not clear how naturalistic mechanisms could produce organisms who are capable of reason and free will.
• 40:43 Does naturalistic evolution have an answer for conscious minds, reasoning, free will? No, consider the work of atheist scholar Thomas Nagel, who argues in his book “Mind and Cosmos” (Oxford University Press 2012) that the existence of mind is a disproof of the neo-Darwinian explanation for life. Darwinism stops us from accepting the reality of minds.
• 42:06 So do naturalistic evolutionists have to explain away the mind as an illusion? First, we humans have immediate experience of consciousness, reason and free will. Second, our whole legal system is based on the idea free will, because you can’t hold someone guilty unless they chose to do something they knew was wrong. Third, we have an epidemic of suicide among young people. This is caused by a crisis of meaning. Intelligent design opens up the possibility of their being a mind behind the universe, who we could have a relationships with.
• 44:53 Why aren’t schools allowed to be honest about the problems with neo-Darwinian evolution? The intellignt design view is to that teachers should be allowed to teach all the vidence for Darwinian evolution, and also discuss some of the problems with the theory. Students learning science should not be told that everything is solved. Students learn science better when they are presented with peer-reviewed evidence for and against a theory, rather than being indoctrinated.
• 47:37 Is intelligent design theory connected to God? Intelligent design infers from the information content in nature that a mind with capabilities like ours injected information into living systems. Intelligent design is agnostic about the designer, because in principle, embodied or unembodied agents could inject information into living systems. Intelligent design is friendly to theism, because theists will immediately identify the mind as God. Furthermore, the fine-tuning in the initial conditions of the universe is another intelligent design argument. In that case, since the design occurs at the beginning of the universe, the intelligent agent acting prior to the creation of the universe would have to be supernatural, i.e. – God.
• 50:53 Can naturalists say that the imposition of “function” on a sequence is arbitrary, in the same way that the English language is arbitrary? This won’t work, because biological function is not arbitrary in the same way as language. Biological function is not arbitrary, because sequences can be tested for function objectively by observing whether sequences can perform functions necessary for life, e.g. – replication.
• 52:43 Doesn’t the multiverse explain away the improbabilities of the fine-tuning, the origin of life, and the development of life? No, because all models of the multiverse require fine-tuning in the mechanism that generates the different universes.
• 55:42 What about cosmological models that eliminate the beginning of the universe? The standard Big Bang model and the inflationary model both posit a beginning of the universe. There is also the Borde-Guth-Vilenkin theory which proves that any universe that is expanding requires a beginning. The only chance for naturalists is quantum cosmologies, but this doesn’t work because 1) it requires an abstract reality of mathematics to actualize the physical universe, but this presupposes a mind. 2)  The model requires an earlier input of information, which can only have come from a mind.

If you listen to this lecture, or watch the video, and you enjoyed it, then please share. If you have questions, I can try to answer them for you – just leave your question in the comments, or shoot me an e-mail, or message me from the blog’s Facebook page, or direct message me on Twitter. I’ve invested a lot of time in this, and have been studying intelligent design since the late 1990s.