Pushing a loaded wheelbarrow
© Matteo / CC BY
Pushing a loaded wheelbarrow across a flat driveway.
In previous articles, I demonstrated how substantial quantities of biological information cannot emerge through any natural process (see here and here), and I described how such information points to intelligent design. Now, I am addressing the mistakes typically made by critics who challenge these claims (see here, here, here, and here). See my post yesterday, here, on misapplying information theory.

A second category of errors relates to arguments against the conclusion that the information content of many proteins is vastly greater than what any undirected process could generate. Most of the critiques are aimed at the research of Douglas Axe that estimated the rarity of amino acid sequences corresponding to a section of a functional β-lactamase enzyme. Many of the attacks result from the skeptics' failure to properly understand Axe's 2004 article in the Journal of Molecular Biology or the underlying science.

Random Processes

The most common mistake is to appeal to studies that demonstrate that random processes can generate structures that perform very simple functions. For instance, our immunity system can manufacture at least a trillion unique antibodies, and at least one will typically bind to any invading germ. This achievement is possible since the probability is relatively high for a random search to locate an amino acid sequence that sticks to some molecule, so the required amount of new information is relatively small. For instance, only a few billion trials are needed to find an antibody that can bind to an antibiotic molecule and break it apart. The problem is that this task is much easier than randomly generating an entirely new amino acid sequence that folds into an enzyme's three-dimensional structure and performs the required complex structural (conformational) changes. Highly specified dynamic structures are required to support an enzyme's often very complex chemical activities.

In contrast, antibodies are comprised of unchanging constant regions that already provide the needed structural support, so only generating functional variable regions is required. The difference between a random search finding an operative variable region and a random search stumbling upon a novel functional enzyme is comparable to the difference in difficulty of pushing a loaded wheelbarrow across a flat driveway and pushing it across a tightrope suspended across Niagara Falls.

The same general misunderstanding holds true for research presented by critics on polypeptides and DNA/RNA sequences. For instance, experiments have generated libraries of randomized amino acid (polypeptide) sequences, some of which could bind to ATP and, after rounds of selection, accelerate its breakdown. Yet, the generated chains do not have the same characteristics as natural proteins. As Axe stated in his 2004 study, the focus of his research was on true enzymes which performed
...not mere catalytic activity but rather catalysis that is mechanistically enzyme-like, requiring an active site with definite geometry (at least during chemical conversion) by which particular side-chains make specific contributions to the overall catalytic process.
The polypeptides do not meet these criteria. Unlike true enzymes, they simply bind to the ATP molecule and break it down into phosphate and ADP, but they do not release the ADP and then repeat the reaction. Research on small nucleotide sequences has also only generated RNAs that perform the simplest of tasks not in any way comparable to true enzymes.

Smoke and Mirrors

In the past, the general public lacked the technical knowledge to decipher the science underlying the evidence for protein rarity, so they were powerless to see past the critics' smoke and mirrors (see here, here, and here). Fortunately, a straightforward analysis of the research by protein expert Dan Tawfik (see here, here, and here) not only confirms and generalizes Axe's results, but is much more accessible to the public. Tawfik's research on β-lactamase yielded results that almost perfectly confirm Axe's rarity estimate. In addition, the former's research and research on the HisA enzyme demonstrate that randomly altering less than 2 percent of the enzymes' amino acids disables them over half of the time. And, altering 10 percent will disable them nearly 100 percent of the time. In contrast, altering 2 percent of a paragraph written in English is usually barely noticeable, and altering 10 percent still leaves a paragraph largely readable. Therefore, protein sequences are often far rarer than readable English sentences, so they are even more difficult to generate by chance.

Tomorrow: Misunderstanding the creative potential of evolutionary processes.