The 2014 PNAS paper, “The predictability of molecular evolution during functional innovation,” presents a large-scale laboratory experiment with E. coli bacteria. By deleting essential genes and watching the bacteria adapt, the authors claim to have created a “predictive framework for the molecular basis of evolutionary innovation.” Their findings are presented as a window into how new functions arise, a process central to the grand narrative of molecules-to-man evolution.
However, a critical analysis of their methods and results reveals the opposite. The study does not provide evidence for the origin of new specified information or functional systems. Instead, it serves as a powerful, large-scale demonstration of the profound limits of random mutation and natural selection. The observed adaptations are not genuine innovations but desperate, degenerative trade-offs made possible by tinkering with pre-existing, brilliantly engineered systems. The evidence, when stripped of its evolutionary gloss, points not to the creative power of unguided processes, but to the designed robustness of an original masterpiece now subject to decay.
A Fair Summary of the Research
The researchers’ goal was to understand the types of genetic changes that allow an organism to overcome the loss of an essential function. They began with 87 different strains of E. coli, each engineered with a single gene deleted from its genome. Each deletion rendered the bacteria unable to grow on a minimal medium, as the missing gene was essential for a specific metabolic task.
They then subjected these disabled strains to experimental evolution, culturing them for approximately 145 generations to see if any could regain the ability to grow. Their key findings were:
- Limited Success: Of the 435 initial populations (five replicates for each of the 87 deletion strains), only 68 managed to evolve a compensatory solution. The vast majority of the bacterial populations (over 84%) went extinct or failed to adapt. Successful recovery only occurred for 22 of the 87 types of gene deletions.
- Two Types of Solutions: By sequencing the genomes of the surviving bacteria, the authors identified two main classes of adaptive mutations:
- Structural mutations: These were typically single amino-acid changes in existing proteins, altering their physical structure.
- Regulatory mutations: These were changes in the non-coding regions of DNA that control how much of a protein is made (e.g., mutations in gene promoters).
- Predictable Patterns: The type of solution that evolved was not random. When the deleted gene was part of a “building block biosynthesis” pathway, the bacteria tended to evolve regulatory mutations. Furthermore, these regulatory mutations almost always affected genes that were functionally “close” in the metabolic network to the function of the gene that was lost. In contrast, structural mutations tended to affect proteins in distant, unrelated pathways.
- Mechanism of Adaptation: The regulatory mutations typically worked by making simple changes, like a single base-pair substitution, that moderately increased the expression of an existing gene. The structural mutations involved slightly altering an existing protein, presumably allowing it to perform a new, compensatory task inefficiently.
The authors concluded that these patterns provide a “predictive framework” for how evolution creates novel functions, suggesting that the context of a gene within the cell’s network determines the path that evolution will take.
The Core Analysis: Adaptive Degeneration, Not Innovation
The paper’s entire premise rests on a semantic sleight of hand: equating “compensation” with “innovation.” The experiments do not show the origin of anything genuinely new. They are a case study in how a complex system responds to being broken. This is not the stuff of molecules-to-man evolution; it is a clear illustration of what Dr. Michael Behe calls the “first rule of adaptive evolution”: break or blunt an existing gene if doing so provides a short-term survival advantage.
1. The “Assume a Gene” Fallacy and the Unsolved Information Crisis
The central problem for evolutionary theory is the origin of the specified information required to build novel genes, proteins, and body plans. This study completely sidesteps this problem. The experiments begin with E. coli, an organism already possessing a genome packed with thousands of information-rich genes and complex regulatory systems. The observed “innovations” are nothing more than minor modifications—a single amino acid tweak here, a promoter mutation there—to this vast, pre-existing library of information.
The authors did not witness the origin of a new protein fold, which Dr. Douglas Axe has shown to be a combinatorial problem of such staggering improbability (a 1 in 10^77 chance for a modest protein) that it is beyond the reach of random mutation in the entire history of the universe. They did not see the origin of a new gene regulatory network. They saw a bacterium, with its key broken, hot-wire its own ignition. To call this “innovation” is to fundamentally misunderstand the challenge. The study explains the modification of existing information, not its origin, thereby failing to address the core “signature in the cell.”
2. Regulatory “Innovation” Is Just Turning Up the Volume
The study found that regulatory mutations were a common solution. However, these were not the sophisticated changes needed to build new developmental programs. They were simple point mutations that caused an existing gene to be over-expressed. This is analogous to fixing a dim light bulb by dangerously overloading the circuit—it might provide more light for a short time, but it is a crude, costly, and ultimately degenerative solution.
This over-expression of a promiscuous enzyme (one with a secondary, weak function) allows it to partially compensate for the lost function. This is not a creative act of evolution; it is a testament to the built-in, designed robustness of the organism. An intelligent engineer would design systems with overlapping functionalities and failsafes. The bacteria are simply activating a pre-existing backup system.
3. The Vast Landscape of Failure
Perhaps the most telling result is the one the authors downplay: for 65 of the 87 gene deletions (75%), evolution was completely powerless to find a solution. The bacteria simply died. Even in the cases where recovery was possible, the majority of the replicate populations went extinct.
This demonstrates that the functional landscape is not a gentle, smooth hill that can be easily climbed by random mutations. It is a treacherous terrain of isolated functional peaks separated by vast, uncrossable valleys of non-function. The fact that evolution failed over 84% of the time, even when dealing with a single gene deletion in a simple organism, exposes the impotence of the mutation/selection mechanism as a creative force. It cannot build new systems; it can barely patch existing ones.
4. The Predictable Pattern Is a Signature of Design, Not Chance
The authors celebrate the “predictability” of the solutions as a victory for evolutionary theory. But this is precisely the opposite of what one would expect from a truly random, unguided process. A blind search would be unpredictable.
The fact that deletions in biosynthetic pathways are preferentially compensated by regulatory tweaks to functionally-related genes is a hallmark of intelligent, modular design. Any competent engineer builds local repair and feedback mechanisms into a complex system. The observed pattern is not evidence of evolution’s foresight, but of the original Designer’s. The organism is activating a pre-programmed adaptive response, a capacity best explained by the Nonrandom Evolutionary Hypothesis (NREH), where organisms are front-loaded with systems to generate targeted, adaptive changes in response to environmental stress.
The Alternative Explanation: Engineered Robustness and Inevitable Decay
A far better explanation for these results is found within a framework of creation and decay. The E. coli bacterium is a marvel of nano-engineering, created with an information-rich genome and designed with the capacity for robust adaptation to changing environments.
- Front-Loaded Information: The thousands of genes, the promiscuous enzymes, and the complex regulatory networks are not the products of chance but are features of an original design. This pre-existing information provides the raw material for the organism to adapt.
- Designed Adaptive Capacity: The predictable patterns of compensation are not lucky accidents but the triggering of built-in adaptive subroutines. The organism was engineered to respond to specific types of damage in specific, targeted ways, explaining why regulatory changes were linked to biosynthetic pathways.
- Genetic Entropy: The “solutions” found by the bacteria are not steps up the evolutionary ladder; they are steps down. Over-expressing a gene or forcing a protein to perform a task for which it was not optimized comes at a pleiotropic cost, reducing overall fitness and leading to what is best described as “adaptive degeneration.” The study’s observation that many recovered clones had longer lag times supports this. The experiment is a perfect microcosm of Dr. John Sanford’s principle of Genetic Entropy: all complex information systems, including genomes, are in a state of inexorable decay from a past state of higher perfection. Adaptation is largely the process of managing this decay.
Conclusion
The study by Blanka et al., while an impressive feat of experimental microbiology, ultimately fails to support the grand claims of Darwinian evolution. It provides zero evidence for the origin of the specified biological information necessary to build new functional genes and proteins. The observed “innovations” are merely degenerative workarounds—crude patches applied to a pre-existing, exquisitely designed system.
The overwhelming rate of failure demonstrates the impotence of random mutation, while the predictable patterns of success point not to the power of a blind search, but to the foresight of an intelligent engineer. The evidence is far more consistent with a model of a recently created, information-rich organism designed with the capacity for limited, pre-programmed adaptation, which is now undergoing a process of decay. This experiment does not show us how evolution builds; it shows us how a designed system breaks.
Leave a Reply