Scientists largely ignored the mathematicians who pointed out the seeming impossibility of creating new genes and proteins because the mathematical equations of population genetics pointed to a nearly limitless creative power of random mutations. Given known mutation rates, population sizes, and reproduction rates, there seemed to be no end to what evolution could accomplish. The problem with this conclusion is that population genetics is based on some assumptions that we now know to be false.
When the neo-Darwinian synthesis (the idea that evolutionary change is driven by natural selection acting on random genetic mutations) was formulated in the 1930s, biologists did not yet understand the structure of genes. Watson and Crick would not discover the structure of DNA and the digital code it uses to build proteins until 1953 and beyond. Prior to this, genes were only understood functionally as those entities that determine visible and selectable traits such as eye color and the number of toes on our feet. It was assumed that single genetic mutations could alter genes in such a way so as to produce a new function, and that one gene could be responsible for building a complex structure. Given these assumptions, it’s easy to envision an organism slowly improving one mutation at a time. Today, however, we know that these assumptions are patently false:
- Hundreds of proteins are often required to create a complex system. To build that system would require changes to hundreds of genes. Furthermore, because the function of these systems depends on the coordination of several well-matched parts, these new proteins (or proteins with new functions) must arise at the same time.
- To change or improve the function of a single protein typically requires multiple, coordinated amino acid changes, which in turn requires multiple, coordinated changes in the DNA. Because the new function depends on the coordination of multiple mutations, every mutation must be present in the protein at once.
For example, ligand binding sites on proteins are necessary for molecules to bind to proteins to form larger, more complex functions. These binding sites, however, require several specific amino acids. All necessary amino acids would have to arise together before there could be a new function, which requires multiple coordinated mutations. Also, protein interactions with other proteins require the matching of several amino acids. If the right combination is not present, there is no function.
If population genetics is going to tell us how much evolutionary change is possible in organisms, it must take into account the probability of multiple, coordinated mutations occurring within a gene, and the probability that multiple, coordinated genes would be produced by natural selection working on random mutations. When the equations of population genetics are updated to account for what we know about what is required to create novel biological function and systems, the odds that any given species could evolve into a new species are prohibitively low.
To see the problem, consider the lottery. If you only have to guess a single number in a bin containing 50 numbers to win a prize, the odds of guessing the right number are 1 in 50. But if you have to guess two numbers to obtain the prize, your odds diminish to 1 in 2,450. If you have to guess three numbers correctly to obtain the prize, your odds diminish further to 1 in 117,600. If you have to guess five numbers correctly to obtain the prize, your odds plummet to 1 in 254,251,200. As you can see, with each additional number required to win the prize (coordinated numbers), the odds of guessing the right numbers decrease exponentially. The same is true in biology. With each additional mutation required to produce a new function, you decrease the odds of producing that new function exponentially. Population genetics has always assumed that the evolutionary lottery was won just by guessing a single correct number over and over again over time. But now we know that many, if not most evolutionary advancements require nature to properly guess multiple numbers at the same time, which lowers the odds of creating new biological function exponentially.
Of course, one could still ask whether the odds can be met. What are the odds that someone will win that lottery? It depends on two things: the frequency of the drawings, the number of players. If there were only 1 million players and there was only one drawing per year, then it would take 254 years before someone would guess the right numbers by chance alone. On the other hand, if there were 254 drawings per year, odds are that one of those one million players would win the lottery once each year. Similarly, to determine whether or not the biological odds can be met depends on the mutation rate, population size, and reproduction rate of an organism. The larger the population size and the faster the mutation and reproduction rates, the greater the odds that an organism will be able to stumble on the right combination by chance alone within the timeframes required by evolutionary theory. Conversely, the smaller the population size, and the slower the mutation and reproduction rates, the odds that an organism will be able to stumble on the right combination by chance alone within the timeframes required by evolutionary theory diminish.
So what are the odds of an organism stumbling on multiple, coordinated mutations required to produce even the tiniest functional change? Virtually zero. Michael Behe argued that if just two coordinated mutations are required, it would require 1 million generations in a population of 1 trillion or more multicellular organisms. But that population size is larger than all the species of multicellular life that exist on this planet at any given time. If we reduce the population size to something more realistic, such as 1 million multicellular organisms, then the number of generations required for evolution to stumble on the right combination of mutations increases to 10 billion. Even if each organism produced a new generation once per year, it would take 10 billion years to get the right combination, which is more than double the age of the Earth and nearly triple the age of life!
The only way for evolution to stumble on the right combination in less than 4 billion years is if the population size is 1 billion organisms persisting over 100 million generations. But this only gets you one new gene function. For the large scale changes required by macroevolution, thousands upon thousands of new genes would be required, and in exponentially shorter time frames. If it takes 3 billion years for nature to stumble on just one double-point mutation to create just one new gene function in a single organism, how in the world could nature produce the huge amounts of novel biological information and function in millions of species in such short periods of time? And if three or more coordinated mutations are required, there isn’t a population size big enough, or generation and mutation rates fast enough to stumble on that combination in under 4 billion years. And remember, we’re only talking about the time it would take to transform the function of an existing gene so that it has a different function by making just two changes to the existing gene. We’re not even talking about building an entirely new functional gene from scratch! When it comes to Darwinian evolution’s ability to explain what we now know about what’s required to generate novel biological function, the math simply doesn’t add up. And that’s a big problem. When the mathematical expression of neo-Darwinian theory – population genetics – shows that neo-Darwinian mechanism are not viable to produce the macroevolutionary changes required by the theory, the gig is up.
Do Darwinists have a recourse? Yes. They appeal to co-option. In this model of protein evolution, a protein that performs one function is co-opted by another evolving system to perform a different function in that system due to some single point mutations that gives it a slightly different function that can be of value to the new system. Douglas Axe and Ann Gauger tested this thesis by finding two proteins that have very similar DNA sequences, and yet perform different functions. They wanted to see if they could transform one protein to do the function of the other by making single point mutations. If single point mutations could not transform the function of these highly concordant proteins, then doing so would require multiple, coordinated mutations, and we are back to the same problem again. Axe and Gauger used proteins Kbl2 and BioF2 for their experiment. They mutated those sites most likely to change the function of Kbl2 into the function of BioF2, but were unable to transform one protein into the other with single point mutations or even large groups of coordinated mutations. This experiment provides evidence for the conclusion that it is not possible to transform even the most homologous proteins into the other via single point mutations, which would falsify the co-option theory of protein development.
Any system whose function depends on the coordinated action of many parts could not be built through the gradual, step-by-step process envisioned by Darwin. Anatomical changes and new biological functions require multiple, coordinated changes, and thus cannot be created via Darwinian mechanisms.
Darwinism can explain the survival of the fittest just fine, but struggles to explain the arrival of the fittest. How do novel biological changes arise? In particular, how do new body plans arise? Body plans can be very different from one another, and a smooth transition from one to the other seems virtually impossible if the organism has to continue functioning during the transition process.
“To create significant changes in the forms of animals requires attention to timing.” Genetic changes that affect processes early in the development of an organism will have a bigger impact than genetic changes that affect processes later on. It’s like building a house. Changes to the foundation as it is being laid down will radically alter the original blueprint for the house, whereas changes to the carpet color – which is laid down near the end of the process – have very little impact. For macroevolutionary changes in an organism’s body plan to occur, there needs to be changes to the genes that regulate body-plan construction very early in an organism’s development. Furthermore, there needs to be changes to the way that proteins are arranged because that will determine how biological systems are built and function (it’s not enough just to generate new proteins). What do we know about what happens when such mutations occur?
In the early 1980s, Christiane Nusslein-Volhard and Eric Wieschaus did a bunch of mutagenesis experiments on fruit flies. They mutated everything that could be mutated, including the genes that code for the development of the fly’s body plan. When they mutated the genes that regulate the body-plan of the fruit fly, inevitably the developing fly died. They admitted that this posed a problem for evolutionary theory, because if mutations in regulatory genes that code for body plans always result in deformation and death, it seems that the origin of new body plans can’t be explained by mutations. Other organisms have been studied since then, with the same results. Mutations that affect the early embryological development of an organism are the least likely to be tolerated by the organism, and result in either deformity or death. As geneticist John F. McDonald writes, “Those [genetic] loci that are obviously variable within natural populations do not seem to lie at the basis of many major adaptive changes, while those loci that seemingly do constitute the foundation of many if not most major adaptive changes are not variable within natural populations.” Meyer summarized his point by saying “the kind of mutations we need for major evolutionary change we don’t get; the kind we get we don’t need.”
The problem the Cambrian explosion poses to Darwinism is the sudden appearance of new forms of animal life. How do you build an animal? What sort of changes are required? Darwinism has to explain the origin of new animal forms in terms of the accumulation of minor changes over time. But now that we know how body plans are constructed and the types of changes required to transform one body plan into another, Darwin’s explanation of how life diversified into so many different phyla is inadequate.
To build a new body plan, not only do you need changes in early-acting regulatory genes, but you also need to change the developmental gene regulatory networks (DGRNs). DGRNs are like the conductors of the genome, determining what parts of the genome get expressed and when.
All cells contain the same genomic information. When an embryo is formed, in its initial stages it just keeps duplicating the same kind of generic cells (stem cells). But at a certain point in early development, these cells are transformed into specialized cells that will form different parts of the body. How is it, if every cell contains the entire genome, that one cell becomes a skin cell and another cell becomes a heart cell? What causes them to differentiate? DGRNs. This is a control system coded into the non-protein-coding sections of DNA (once thought to be “junk DNA”) that regulates which genes within the genome are expressed (content), and when they are expressed (order and timing). These regions of the genome function like circuits, transmitting signals that influence when certain cell types develop. These circuits are highly integrated and resistant to mutational changes because they are structured hierarchically. They are structured like a building, in which you must lay the foundation before the walls and ceiling can be erected. If you change the order in which the pieces are put together, or if you try to remove the foundation, the entire edifice crumbles to the ground. Eric Davidson, a specialist in DGRNs at the California Institute of Technology, explains it this way:
There is always an observable consequence if a dGRN subcircuit is interrupted. Since these consequences are always catastrophically bad, flexibility is minimal, and since the subcircuits are all interconnected, the whole network partakes of the quality that there is only one way for things to work. And indeed, the embryos of each species develop in only one way.
Davidson is very clear that while the neo-Darwinian process can explain small changes in animal populations, it cannot explain the origin of new body plans:
Neo-Darwinian evolution…assumes that all process works the same way, so that evolution of enzymes or flower colors can be used as current proxies for study of evolution of the body plan. It erroneously assumes that change in protein-coding sequence is the basic cause of change in developmental program; and it erroneously assumes that evolutionary change in body-plan morphology occurs by a continuous process. All of these assumptions are basically counterfactual. This cannot be surprising, since the neo-Darwinian synthesis from which these ideas stem was a premolecular biology concoction focused on population genetics and…natural history, neither of which have any direct mechanistic import for the genomic regulatory systems that drive embryonic development of the body plan.
If an organism required, say, 100 new genes to produce a new biological system, and each of those 100 genes arose gradually over time, natural selection would likely weed them out since they served no function (it requires energy to create them, which is wasted if they have no function). If one argued that that many, if not all, served some biological function unrelated to the system they would eventually create, and then, once the 100th gene was created, the other 99 were co-opted from other parts of the organism to create this new system, this would have negative impacts on the other systems those 99 genes were supporting. It’s like Jenga. Imagine having five Jenga towers, and then removing 99 blocks from them (~10 from each tower) all at once in order to construct a new Jenga tower. Such movement would either weaken the structural integrity of the original Jenga towers, or cause some of them to crash.
This is comparable to the difference between changing a few letters in a short poem so that it has a slightly different meaning than the original, as opposed to constructing an entirely new poem with an entirely new meaning from scratch.
Cornell mathematical biologists, Rick Durrett and Deena Schmidt, attempted to refute Behe by doing their own calculations. In their paper, “Waiting for Two Mutations: With Applications to Regulatory Sequence Evolution and the Limits of Darwinian Evolution,” they argued that Behe’s timelines were too long. So what did they find? They found that it would take “only” 216 million years to fix two coordinated mutations in the line of hominids leading to humans. The problem should be evident: even this is more than 30x the amount of time allowed for hominid evolution. If it takes 216 million years to fix just one such mutation, and human evolution would require hundreds, if not thousands, of such mutations, then human evolution is dead in the water. So both critics and supporters of Darwinism alike have demonstrated that evolutionary mechanisms are not sufficient to account for the kind of changes necessary for macroevolution.
Stephen Meyer, Darwin’s Doubt: The Explosive Origin of Animal Life and the Case for Intelligent Design (Harper One: , New York, 2013), 259.
John F. McDonald, “The Molecular Basis of Adaptation: A Critical Review of Relevant Ideas and Observations,” Annual Review of Ecology and Systematics 14 (1983): 77-102, 93.
Stephen Meyer, Darwin’s Doubt: The Explosive Origin of Animal Life and the Case for Intelligent Design (Harper One: , New York, 2013), 262.
Eric Davidson, “Evolutionary Bioscience as Regulatory Systems Biology,” Developmental Biology 357 (2011): 40, quoted in Meyer, 268.
Eric Davidon, “Evolutionary Bioscience as Regulatory Systems Biology,” Developmental Biology 357 (2011): 35-36, quoted in Meyer, 269.