by Clete Knaub
Molecular homology has been acclaimed as the field of study that saved the house of evolution from collapsing by serving as an independent check that confirms evolution to be a fact.1 What is molecular homology? Is it an independent confirmation of evolution? Can it "clock" the course of evolution?
To answer these questions, let's consider first the techniques involved in molecular homology. Basic to the method is the fact that the structure and function of all living organisms depends on biologic molecules called proteins. These proteins are in turn made up of carbon-based molecules called amino acids. Amino acids are the links; proteins are the chains forged from these links. The specific sequencing of amino acids determines the exact nature and function of the protein. For instance, a protein with one specific arrangement of amino acids will serve to digest fat in our stomachs while another sequence specifies that the molecule will carry oxygen.
Except for some mutations (which are virtually always deleterious), the same amino acid sequence for one particular kind of protein is present in all organisms of the same species. Between species, however, the amino acid sequence for a protein such as alpha-hemoglobin, for example, can and usually does vary. Comparing these differences between two or more species and drawing inferences from these comparisons is the field of molecular homology.
Scientists now possess the technology to extract and, with a fair amount of accuracy, determine the amino acid sequences of various proteins. For example, the hemoglobin sequences for man, mouse and horse can be determined. When the sequence of the man's hemoglobin is then lined up with the sequences of the mouse and horse, the total number of amino acid differences between these three species is determined by simply doing a pairwise comparison of each amino acid at every position along the entire length of the molecule.2 Considering a hypothetical example, the total number of differences between man and mouse might be five and for the man and horse seven; then, from the evolutionist's point of view, the man must be more closely related to the mouse than to the horse.
Although evolutionary relationship is hardly the only logical inference from amino acid sequence comparisons, at least it is a potentially reasonable interpretation of the data. However, when this pairwise comparison is strictly followed to produce a phylogenetic (evolutionary) tree, many results embarrassing for evolutionists are obtained. Perhaps the most widely pictured evolutionary tree is based on cytochrome c; yet it shows the turtle is more closely related to the birds that to its fellow reptile, the snake. Furthermore, the chicken is grouped with the penguin rather than the duck, and man and ape separate from the main mammalian branch before the supposedly less advanced marsupial mammal, the kangaroo.3
The cytochrome c tree pictured in books and magazines is only one of forty trees generated by computer analysis of the data—the tree "corrected" for closest fit to the "known phylogeny" (i.e., the presumed evolutionary history).4 Certainly such a tree cannot be claimed as independent confirmation of evolution.
Amino acid sequence data are actually heavily "massaged" even before they are used to construct these evolutionary trees. Since it is mutational changes in DNA that are presumed to produce, ultimately, the differences in amino acid sequences, estimates of silent mutations, one vs. two step changes in codons, several changes at one position, and estimates for other such corrections must be made. When the raw data are actually antigen-antibody tests or DNA hybridization, as often is the case, uncertainty regarding even amino acid differences, let alone amino acid changes, becomes considerable.5 The computer must be told in advance to generate only ancestral sequences that allow for further ancestral sequences,6 otherwise, as we observed in some of our analyses, intermediate sequences are generated that break the presumed evolutionary chain.
In spite of all these problems, an ever-hopeful evolutionist predicted that a more accurate molecular phylogenetic tree for species would be obtained when "additional proteins and nucleic acids have been determined."7 Quite the opposite has taken place since he made that prediction. The more protein sequences determined, the less likely the combined tree represents the accepted classical evolutionary tree. Workers with sequences for LH (luteinizing hormones) were "forced" to postulate that amphibians evolved directly into mammals, instead, of first into reptiles.8 Vincent Demoulin said that the composite tree including data from many different kinds of cytochromes simply "encompasses all the weaknesses of the individual trees."9 Reviewing recent data before a prestigious group at the American Museum Nov. 5, 1981, Colin Patterson, himself an evolutionist, stated that if only the data of molecular homology were considered, then descent from common ancestry, the foundational concept of evolution, was "precisely falsified."10
In view of the failure of data from molecular homology even to support evolution, it is surprising that molecular homology has also been used as an "evolutionary clock," i.e., to try to determine rates at which mutations became fixed in populations. This is done in two basic steps. Step one is to construct a phylogenetic evolutionary tree based on protein homologies. Step two is to determine, from the fossil evidence, when the species diverged from each other. Suppose that the man and mouse in our previous example shared a common ancestor twenty million years ago by evolutionary reckoning. That would be five amino acid differences in twenty million years, or one amino acid changing "accepted point mutation" (PAM) per four million years. If that rate is relatively constant for most proteins, then the calculation can be "reversed" to determine times of divergence from measured (and "massaged") amino acid sequence differences.
Like tree construction, molecular clocking seems an invitingly simple and straightforward use of protein differences, but, like tree construction, the clocking procedure is plagued with a plethora of problems. First, the "ticking" of the clock is invisible—it is presumed "accepted point mutations" or PAMS, which have never been observed, not measured mutation rates. Second, the assumption that rates of evolution should be approximately the same for most proteins is considered absurd and even anti-evolutionary by the classic school of evolutionary thought, the selectionists, and the extreme variability of estimated rates seem to bear out their concern. Speaking of an average rate is somewhat like saying that on the average all animals have the same temperature, a statistical deception that communicates what Patterson might call "anti-knowledge." Third, we have already seen how imperfect are the evolutionary trees constructed on the basis of molecular homology. Finally, evolutionists have at last been forced to admit publicly that fossil evidence contains virtually no transitional forms and that, instead, it often suggests the simultaneous, explosive appearance of diverse types. Therefore, it is at least nonsense to try to determine when species diverged from each other, and it may be worse, since types widely different from each other seem to have diverged at essentially the same time (from unknown ancestors).
In view of all these difficulties, it is not surprising that, in a major review of molecular clocks, Walter Fitch dismissed discussion of the clock dependent on paleontological dates, since, as he put it, any discrepancy could easily be due to an error in the dating. He then turned to a "calibration-free" test of the clock, but had to admit that no satisfactory statistical test of the clock had yet been done.
All these problems in principle crystallize as problems in practice when the clock is applied to the snake/bird/turtle example. Fitch points out, that a relative clock distance of 55 between snake and turtle and only 11 between bird and turtle (making the two reptiles much less related than the bird/turtle pair) is, of course, a considerable distortion of our current biological viewpoint."11 Fitch's explanation for this conflict between fact and evolutionary theory should jolt evolutionists and scientists alike: " … the truth really is that either one or more of the sequences is incorrect, that our current view of amniote phylogeny [land animal evolution] is incorrect, or both." Earlier, Fitch had pointed out a third general source of error: the divergence times could be in error. Sadly, Fitch does not acknowledge a fourth source of error, namely, failure of the molecular clock itself. Presenting evidence that would impress a skeptic only as an "escape from reason," Fitch clings to the clock hypothesis, even after showing that the springs and cogs of the mechanism are scattered and broken.
There is so much "slop" in both the data and its processing in the field of molecular homology that Colin Patterson, senior paleontologist at the British Museum, spoke of it as "anti-knowledge" generating "anti-theory," apparently meaning a false assessment of the facts inducing the false concept that evolutionary common ancestry offers some sort of explanation for the data (which, as he points out, has already been "massaged with evolutionary theory").12 In fact, after pointing out that the current evolutionary explanation for molecular homology was "precisely falsified," he went on to consider a creationist explanation for the data.
Lest evolutionists try to divert attention from weaknesses in their view by ridiculing another view (Macbeth's "best-in-field fallacy"13)—Let us content ourselves first to establish the demise of one theory and the need for another. After all, good scientists who know that neither genetics nor paleontology suggest evolution have been seduced into believing that the rickety edifice of evolution has somehow been shored up by the enthusiastic—but entirely baseless—claims of the self-seduced molecular evolutionists.
1. Klassen, G., "Scientific Creationism vs. Evolution," Mennonite Mirror, Oct. 1981. (Glen Klassen is a member of the Dept. of Microbiology at the University of Manitoba.)
2. Fitch, W., and E. Margoliash, "Construction of Phylogenetic Trees." Science, V. 155, 1967, p. 279.
3 . Ayala, F., "The Mechanisms of Evolution." Scientific American, V. 239, No. 3,1978, p. 56.
4 . Fitch, W., and E. Margoliash, loc. cit., p. 281.
5. Fitch, W., Molecular Evolution, (F. Ayala, editor), Sinauer Associates, 1976, p. 160.
6. Dayhoff, M., "Atlas of Protein Sequence and Structure," V. 5, No. 3, 1978, p. 345.
7. Ayala, F., loc. cit., p. 68.
8. King, J., and R. Millar, "Heterogeneity of Vertebrate Luteinizing Hormone-Releasing Hormone," Science, V. 206, 1979, p. 67.
9. Demoulin, V., "Protein and Nucleic Acid Sequence Data and Phylogeny." Science, V. 205, 1979, p. 1036.
10. Patterson, C., as quoted by Sunderland, L. and G. Parker, "Evolution? Prominent Scientist Reconsiders." Impact No. 108, Institute for Creation Research, 1982.
11. Fitch, W., loc. cit., p. 174.
12. Patterson, C., loc. cit.
13. Macbeth, N., Darwin Retried, Gambit, 1971.
Authors: * Graduate student in Biology, ICR Graduate School. ** Professor of Biology, ICR Graduate School.
Cite this article: Clete Knaub. 1982. Molecular Evolution?. Acts & Facts. 11 (11).