If a little is good, more must be better, right? Though it is tempting to live by this adage, it seldom turns out to be true. More is rarely better.
Evolutionary biologists have begrudgingly come to the same conclusion recently when it comes to building evolutionary trees using DNA sequences. More sequences are not better.1 This vexing problem for the evolutionary paradigm encourages those who seek alternative explanations for life’s history.
Building Evolutionary Trees
If biological evolution explains life’s origin and history, then many scientists believe it should be possible to reconstruct this history and depict it in the form of a tree—a diagram that describes evolutionary relationships among all organisms as nested branches that all root in a common ancestor.
Traditionally, evolutionary biologists built these trees (called phylogenies) using anatomical features—physical similarities and differences. A few decades ago, as the technology emerged to sequence DNA (the molecule that harbors genetic information) evolutionary biologists increasingly focused their efforts on using DNA sequences to construct evolutionary trees (called molecular phylogenies).
This shift was driven by the belief that DNA sequences—a direct expression of inheritance—contain much more foundational information about evolutionary relationships than anatomical features. Still, researchers predicted that molecular phylogenies should agree with trees built from anatomical data (morphological phylogenies––or, types of “tree branches”). They also predicted that molecular phylogenies constructed with DNA sequences from one region of the genome should agree with evolutionary trees built from other genomic regions.
As I have pointed out previously, neither prediction holds true. It is commonplace for molecular phylogenies to contradict morphological phylogenies and one molecular phylogeny to disagree with another.2
In my view, this incongruence represents a significant problem, prompting justified skepticism about the validity of the evolutionary paradigm to fully account for life’s history and diversity. Undeterred, biologists predicted harmony would be achieved among molecular phylogenies once they gained access to larger stretches of DNA sequences.
New Hope Rooted in Genomics
Evolutionary biologists have gotten their wish with the onset of a new era in the life sciences: the genomics age. Researchers now have ready access to large stretches of DNA sequences. In fact, they can routinely sequence and analyze the entire genome of an organism. And it is vogue for biologists to attempt to build evolutionary trees using genome sequences (called phylogenomics).
More Is Not Better
Are the evolutionary trees looking healthy? No. More DNA sequence data is not better. It actually makes things worse, as a recent analysis highlights.3 The use of super sequences of DNA to build evolutionary trees actually confirms the problem by generating conflicting molecular phylogenies that display statistically robust incongruencies. (Also see Fazale ‘Fuz’ R. Rana, “Origin of Complex Cells: A Big Event for Evolution or Creation?” New Reasons to Believe Vol. 3, No. 1 (2011), 9–10.)
A team of researchers thinks they’ve identified the glitch: phylogenetically noninformative DNA sequences intermingled among those that can be used to build reliable evolutionary trees. The researchers propose some reasons why DNA sequences can lose information about their putative evolutionary history. As they point out, simply adding more DNA sequences to the tree-building process adds both informative and noninformative sequences.
Their solution: eliminate noninformative sequences from the analyses. This solution is not as easy as it seems, however. Deciding which DNA sequences to remove seems largely arbitrary. How does one truly know whether or not a sequence is confounding the analysis? How does one know for certain that a suspect sequence doesn’t actually harbor valuable information?
The answer: nobody knows. No objective criteria exist that can be used to distinguish between useful DNA sequences and those that frustrate the tree-building enterprise. A sequence is deemed to be phylogenetically noninformative, after the fact, if it yields an undesired evolutionary tree. And how does one know that an evolutionary tree is undesired? Because it depicts unexpected evolutionary relationships. This proposal smacks of circular reasoning.
Incongruence and the Creation Model
Could it be possible that the incongruence reflects a rudimentary problem with the evolutionary paradigm? In science, contradictory results are a telltale indicator that something is fundamentally wrong—with the data, the methodologies, or the theory.
This disagreement with one paradigm can be taken as evidence for another. According to the late evolutionary biologist Morris Goodman:
If the biblical account of creation were true, then independent features of morphology, proteins, and DNA sequences would not be expected to be congruent with each other. Chaotic patterns, with different proteins and different DNA sequences failing to indicate any consistent set of species relationships, would contradict the theory of evolution.4
- Hervé Philippe et al., “Resolving Difficult Phylogenetic Questions: Why More Sequences Are Not Enough,” PLoS Biology vol. 9, no. 3 (March 2011): e1000602. Doi: 10.1371/journal.pbio.1000602.
- Henry Gee, “Ending Incongruence,” Nature 425 (2003): 782.
- Olivier Jeffroy et al., “Phylogenomics: The Beginning of Incongruence?” TRENDS in Genetics 22 (2006): 225–31; Hervé Philippe et al., Resolving Difficult Phylogenetic Questions
- Morris Goodman, “Reconstructing Human Evolution from Proteins,” in The Cambridge Encyclopedia of Human Evolution, Steve Jones, Robert Martin, and David Pilbeam, eds, (New York: Cambridge University
Press, 1992), 307–13.