Too Good to Be True: Evolution and the Origin of Bioinformation
New work by researchers from the Georgia Institute of Technology impacts one of the more popular arguments used to support the creation/intelligent design perspective: namely that it would be highly improbable for the vast amount of genetic information housed in DNA to arise through chance events. This research exposes a flaw in this argument, but I present a separate study that bolsters the creation view in a way that avoids common pitfalls.
Anyone who has received an email from Nigeria promising great wealth knows that if an offer appears to be too good to be true, it probably is.
A study on protein structure by researchers from Georgia Tech demonstrates that this cautionary principle also applies to the arguments popularized by creationists and intelligent design advocates.1 A favorite approach of Christian apologists involves emphasizing the vast amount of information contained in biomolecules—such as the nucleic acids (DNA and RNA) and proteins—and the astronomical improbability of chance events ever generating that content. They conclude that biochemical information and life must originate from a Creator. Yet, the Georgia Tech team’s research makes it look like chance processes can generate functional proteins.
The group investigated the origin of binding interactions between small molecules and proteins. These interactions are central to proteins’ function in the cell and usually display a high degree of specificity. Small-molecule binding takes place in pockets (concave regions) at the protein surface. Precise interactions between the small molecules and the chemical groups that line the pockets establish the binding specificity and make it possible for a protein to discriminate among the vast number of small molecules in the cellular environment.
The researchers compared the binding pockets for all the known proteins found in nature (housed in the Protein Data Bank) with the concave surface structures of artificial protein-like materials found in a database of randomly generated structures. They discovered that only around 400 different types of binding pockets exist in naturally occurring proteins. Remarkably, all of these binding pockets also occur in the randomly generated database of artificial protein-like materials. It looks as if highly similar pockets occur for a variety of protein structures and that a protein’s ability to bind small molecules results simply from its structural and physicochemical properties.
Given that such a limited number of binding pockets exist, proteins generated by combining amino acids at random will easily possess the full suite of binding pockets found throughout nature. The Georgia Tech team speculates that each binding pocket in the artificial protein database will bind a wide range of small molecules. They argue that specificity arises as a secondary feature via minute evolutionary changes to the pockets.
This work has profound implications for Christian apologists and intelligent design proponents who argue that biochemical information cannot arise through chance processes. I, too, have made this very argument in the past and regret doing so. In my experience, it plays well to a lay audience, but has little impact on origin-of-life researchers and biochemists. Scientists resist the argument when it is presented this way because it misrepresents the processes that evolutionary biologists credit with generating information-rich molecules and portrays a lack of understanding about protein structure and function relationships (see “Intelligent Design: The Right Conclusion, but the Wrong Reasons”).
The nature of protein space is key to the study of protein structure and function relationships and the impact these relationships have on the origin of life question. Protein space refers to all possible amino acid sequence combinations that conceivably exist. Of all those sequences, some are functional and some are not. Think of the sequence space like an ocean and the functional sequences within it like islands. The big question then becomes: How many “islands” exist in sequence space and what is the separation among them?
Some biochemists (and Christian apologists) have long thought functional islands were sparse because many proteins do not tolerate changes in amino acid composition (substitutions cause loss in function). In other words, they argued that not only is sequence space lightly populated, but also the islands appear to feature steep cliffs.
However, recent work demonstrates that while the islands of functionality do possess steep cliffs, the ocean is richly populated with functional regions. Some biochemists believe that a large part of sequence space consists of a dense meshwork of functionality that connects the islands. The Georgia Tech team’s work provides yet more evidence that protein space may be more densely populated than previously thought. And, as functional regions populate more protein space, the easier it is, in principle, for information-rich molecules to arise via natural processes.
Impact on Apologetics
Does this mean it’s no longer possible to use biomolecules’ information content as evidence for the necessity of a Creator in bringing life into existence? Not at all. It simply means that we must be flexible enough to consider a new approach.
For example, the evolutionary mechanisms’ capacity to generate biochemical information is not without genuine challenges. Recent work by astrobiologist Sara Walker and physicist Paul Davies highlights a fault in attributing biochemical information to naturalistic processes.2
Walker and Davies describe the information content of individual biomolecules as trivial information that, according to their view, is readily produced by evolutionary mechanisms. Rather, they focus on biochemical information’s algorithmic nature. They point out that life is composed of matter that harbors information, which in turn instructs the matter how to behave. This property is referred to as algorithmic information. It has causal efficacy, which means that the information harbored in the biomolecules exerts control over the biochemical systems that these molecules comprise. According to Walker and Davies, though chemical and natural selection can generate the information harbored in an individual protein or gene, there is currently no known mechanism that can account for the algorithmic nature of biochemical information (see “Does New Approach Solve Origin-of-Life Problem?”).
The bottom line is that the information content in biochemical systems still serves as potent evidence that life stems from the work of a Mind, but we must be careful and thoughtful about how we construct the case for the Creator.
For tips on how to make a powerful case using biochemical information, see my previous article: “Intelligent Design: The Right Conclusion, but the Wrong Reasons.”
1. Jeffrey Skolnick and Mu Gao, “Interplay of Physics and Evolution in the Likely Origin of Protein Biochemical Function,” Proceedings of the National Academy of Sciences, USA 110 (June 4, 2013): 9344–49.
2. Sara Imari Walker and Paul C. W. Davies, “The Algorithmic Origins of Life,” Journal of the Royal Society Interface 10 (February 6, 2013): doi: 10.1098/rsif.2012.0869.