Variety Is the Spice of Life

True understanding of the complexity of biological systems demands an assortment of model systems.

By H. Steven Wiley | February 1, 2013


A fundamental goal of science is to discover rules. If we find that two chemicals react to yield a specific compound, then we expect that to reflect a general truth that other scientists can repeat. The situation is not so simple in biology, however, because our experimental results depend, to some degree, on the model system we are using.

Model systems have been around for as long as biology has been a science. We posit a general truth by finding organisms in which that truth is most evident. Darwin used pigeons to study the heritability of biological variation. Mendel used peas to show trait inheritance. Researchers assumed that principles observed in one organism would be true for all, and for a long time, this generally proved true.

For most of the last century, there were nearly as many model organisms as there were scientists. I was lucky to be trained in a laboratory that used all sorts. I studied frogs and toads and goldfish as well as sea urchins and mice, all in the search for mechanisms that controlled ovulation. And why not? After all, estrogens work as well in frogs
as in humans.

Researchers assumed that principles observed in one organism would be true for all, and for a long time,
this generally proved true.

The great stories of biology were stories of different model systems. Coated pits, the fundamental structures of endocytosis, were discovered in mosquito oocytes. We studied their hormonal regulation in frog oocytes. Nobel laureates Michael Brown and Joseph Goldstein investigated them in human fibroblasts. Fundamental cell cycle control mechanisms were worked out using surf clams, frog eggs, multiple yeast species, and cultured human cells. By looking at a variety of systems, general truths unfolded.

The golden age of model-agnostic biology came to an end as people started asking more specific questions. Biological systems are complex, and to understand a process at a mechanistic level, you need myriad details—and details do vary between organisms. People started focusing on systems that were the most convenient and assumed the results would be universally applicable. For example, yeast became ascendant in cell cycle research because of the genetic tools available.

Unfortunately, the more time and effort that biologists invested in particular systems, the more they became wedded to them, regardless of the questions being asked.

In a recent Nature opinion piece, “Model organisms: there’s more to life than rats and flies,” zoologist Jessica Bolker of the University of New Hampshire makes the point that the problem under study should drive the choice of model system (Nature, 491:31-33, 2012). This has always been so, but it is an old lesson that we have forgotten in the wave of reductionism that has swept biology over the last several decades. What holds us back is the accumulated wisdom regarding favored model systems. Because biology, unlike mathematics, lacks a formalized knowledge structure, the most accessible knowledge is in the collective experience of its practitioners. If that experience has been dominated by decades of work in a particular model system, shifting to a new and perhaps more appropriate one can be problematic.

Molecular tools that were formerly restricted to a few favored systems are now widely applicable.

As Bolker correctly argues, focusing on just a handful of model systems inhibits scientific progress. As systems-level research, which integrates multiple types and levels of biological information, becomes more necessary to understand complex diseases, there is an urgent need to find principles that extend across all models, Homo in particular. There is always the danger that a particular mechanism might be a contingent aspect of a particular organism.

Fortunately, the barriers that have traditionally locked us into a particular system are starting to fall. With new high-throughput technologies, one can gather more information in a single experiment than was formerly obtained over years. Molecular tools that were formerly restricted to a few favored systems are now widely applicable.

Using novel and appropriate models to solve long-standing problems is an opportunity for young investigators looking for a competitive advantage. On the grant review panels on which I serve, I am noticing a greater expectation that the model system being proposed is the one most appropriate for the question being asked.

Just as the true value of sequencing the human genome was only realized when a host of other organisms were sequenced, results from a variety of model systems are still necessary to understand the relevance of the favored ones. From differences, we learn the rules. 

H. Steven Wiley is lead biologist for the Environmental Molecular Sciences Laboratory at Pacific Northwest National Laboratory.

Add a Comment

Avatar of: You



Sign In with your LabX Media Group Passport to leave a comment

Not a member? Register Now!

LabX Media Group Passport Logo


Avatar of: John Edser

John Edser

Posts: 24

February 4, 2013


Model use and misuse.


Simplified models can be used or misused to test theories. Different species acting as modelling systems can greatly simplify testing a particular theory but the results must always be corrected for these specific simplifications otherwise the model is misused. It is possible to oversimplify. If we define a modelling simplification as the change/deletion in theory variables, it becomes possible to differentiate an oversimplification as the change/deletion in a theory constant. Of course, an oversimplification is much more serious than just a simplification because theory constants provide a falsifiable frame of reference for a theory allowing empirical separations of cause and effect. Oversimplifications may catastrophically allow a reversal of theory cause and effect. Ending up with no theory constant at all  reduces a theory to just an empty tautology (circular argument) which makes mathematical sense but no scientific sense. 


Unfortunately, simplified generic analysis employed as a model evolutionary theory synthesis has resulted in uncorrected Neo Darwinistic model misuse via the substitution of an organism part conveniently modelled to represent a whole, e.g. the continuing status of the gene as an independent in fitness biological entity. This model was pioneered by Fisher in the early 1900's and developed by Haldane, Hamilton et al and is today popularised by Dawkins "selfish geneism". Early criticism of Fisher's model as misused "bean bag genetics" was ignored.  In the 1960's Waddington attempted to correct some of Haldane's modelling simplifications via including four new variables in Haldane's basic population genetics equations: developed in X, selected in X, developed in Y, selected in Y allowing for the first time, a modicum of genetic epistasis. He was ignored.


Hamilton's Inclusive Fitness model for the evolution of adult organism fitness altruism via independent in fitness "selfish genes" misrepresents Darwinism via the ongoing misuse of a simplified and over simplified model.


Hamilton's proposed rule:




The rule is used to explain the evolution of apparent adult organism fitness altruism in nature via independent in fitness "selfish genes" within poly-centric Neo Darwinism. The cost c, paid by a single donor allowing a group of related r organisms to increase their joint fitness via b has no ceiling and genetic epistasis e, is not included. When e is included (correcting Hamilton's modelling simplification) the rule becomes:




The rule with genetic epistasis included becomes inoperable since relatedness geometrically decreases as the number of epistatically included genes increases from a minimum of 2.


The cost c must be deleted from a missing total that provides a ceiling to c: the total number of adult forms reproduced by the donor  (the donor cannot give away more than this in fitness). It is this total that provides a missing fitness constant, per selectee per population that must be incorporated within the rule otherwise the rule as it stands without a singe constant, only represents a tautologous oversimplification of Darwinism within which Darwinian cause and effect can and has been, routinely reversed.




John Edser


independent researcher







Avatar of: James V. Kohl

James V. Kohl

Posts: 532

February 19, 2013

Reproduction is nutrient-dependent and pheromone-controlled in species from microbes to man. That suggests the molecular mechanisms are the same. Thus, after a thorough review of the extant literature, we can readily conclude that "Olfaction and odor receptors provide a clear evolutionary trail that can be followed from unicellular organisms to insects to humans."

That fact is demonstrable. Substitution of alanine for valine in a human population results in alterations of skin, sweat, and hair. The hypothesis supported is that the enhanced thermoregulation required for evolution at the molecular level fuels adaptive mitochodrial-nuclear interactions. The intranuclear interactions are manifested in phenotypical changes that enable sexual selection for nutrient-dependent reproductive fitness.

Adaptively evolved fitness is signaled by pheromones via the alterations in skin, sweat, and hair in mammals. The problem for some people is the complexity of the systems biology. We are required to get from nutrient fueled energy driven protein synthesis in cells to ecological, social, neurogenic, and socio-cognitive niche construction. That requirement exists to link the sensory environment to protein synthesis via changes in gene expression. The changes in gene expression lead to changes in behavior, and back to changes in gene expression via reproduction.  If not for animal models of that complexity, all hope for understanding would be lost.

The honeybee model organism incorporates what is known about nutrient-dependent pheromone-controlled social behavior. The mouse model extends what is known to sexual selection in mammals.  The physics of thermoregulated DNA strand pairing extends common molecular mechanisms from microbes to man. 

Extending the concept of pheromone-controlled reproduction from microbes to humans is not possible in the current climate of animal model specializations and what are believed to be mutations that somehow cause adaptive evolution. In Drosophila, for example, an experimentally induced valine-alanine point "mutation" reduces fecundity as is consistent with starvation. But "fixation" of the mutation is used to explain adaptive evolution, and diversity in Drosophila is a commonly used animal model.

In my opinion, adaptive mutations should not be used as substitutes for explanations of the epigenetic effects of nutrients and pheromones on epigenesis, epistasis, and adaptive evolution. Instead, differences in perspectives on the  valine alanine variants should be compared to determine how "adaptive mutations" enable nutrient-dependent pheromone-controlled species diversity. What if all adaptations are nutrient-driven and pheromone-controlled?

Let's first get the physics and the biology correct, before we mathematically model the impossible just because it seems that mutations are adaptive at the level of population genetics. Is there an animal model of the biology for that?

Popular Now

  1. Dartmouth Professor Investigated for Sexual Misconduct Retires
  2. Theranos Leaders Indicted For Fraud
    The Nutshell Theranos Leaders Indicted For Fraud

    Federal prosecutors filed criminal charges that allege the company’s promise to revolutionize blood testing swindled investors out of hundreds of millions of dollars and put patients in danger.

  3. Laxative Causes Long-Term Changes to Mouse Microbiome
  4. Probiotics Prevent Cholera in Animal Models