Paper Digest: Addressing Flaws in Population Dynamic Models
Population genetics, the field that studies how alleles shift and spread in a population, can tell us something about our history. A genome functions like a unique fingerprint for every organism, allowing scientists to compare genomes and determine which genomes are similar. According to evolutionary biology, genome comparison can also allow one to trace the gradual accumulation of mutations (as proposed by Darwinian evolution) and reconstruct the history of life through common ancestry.
An ID Perspective
However, intelligent design (ID), as a scientific theory, posits that the origin of certain features of living organisms — including many features of genomes — are better explained by an intelligent cause than solely by undirected processes like random mutations. Leading ID proponents, including Michael Behe, Douglas Axe, Winston Ewert, William Dembski, and Stephen Meyer, have raised concerns about the ability of random mutations to generate complex, information-rich systems. Behe argues that random mutations generally degrade, rather than enhance, genetic information. Additionally, researchers such as Winston Ewert, Ann Gauger, Paul Nelson, Casey Luskin, and others, although agreeing that genome comparison is extremely helpful, have also highlighted issues in phylogenetics, the field that uses genome comparisons to infer common ancestry.
An emerging perspective within the ID community, advanced by figures like Jonathan Bartlett (whose paper I’m covering today), Brian Miller, Steve Laufmann, and myself (here, here, and here), suggests that some genetic variation within a population of organisms represents programmed, parameterized adaptive potential. Bartlett discusses the issue in his recent peer-reviewed article, “Alternate mutation modalities in models of population dynamics” (Academia Biology). He first raises problems with population dynamic models. Bartlett then suggests integrating the parameterized genetic variation into population dynamic models to resolve inconsistencies and improve accuracy, providing a starting example for how this might be accomplished
Key Flaws in Current Models
Bartlett points out that reproductive fitness is often conflated with biophysical fitness. He defines biophysical fitness as the effectiveness of biophysical components to carry out their function. On the other hand, reproductive fitness he defines as “the relative ability of different genetic configurations to produce offspring.”
The problem with this conflation becomes evident when we consider that in two environmental contexts an organism with the exact same biophysical fitness may have different reproductive fitness. One example Bartlett provides is a genetic variation affecting an outer bacterial coat protein. In certain situations, this type of genetic variation can conserve the biophysical fitness of the protein (meaning the protein retains its ability to structurally reinforce the cell wall or outer membrane), even as it reduces reproductive fitness in the environmental context of a specific antibiotic to which the genetic variant makes the bacteria more susceptible. Hence, a genetic change can maintain biophysical fitness but reduce reproductive fitness.
A second problematic assumption is that genetic change is random. Bartlett notes two types of genetic change that challenge this assumption: stress-induced genetic variation and cyclical genetic variation. As an aside, you might have noticed I personally don’t like to call genetic change “mutation” because I think the term “mutation” should be reserved for random genetic change that is known to negatively impact biophysical fitness.
An example of stress-induced genetic variation is that single-celled organisms are known to activate polymerases that introduce genetic variation. This occurs when stress is detected but not in a random pattern. Instead, these polymerases target genetic variation to specific regions of the genome where it might be helpful.
Cyclical genetic variation, also called back mutation, is defined as genetic change which happens on a periodic cycle and may revert to a previous allele. These events happen in phase-variable genes. Bartlett explains one mechanism:
There are multiple mechanisms by which [cyclical genetic variation] occurs, one of which is inverting the promoter sequence. One direction turns the promoter on, and the other direction turns it off. By having a sequence with a bistable state, the organism will revisit both configurations repeatedly.
In summary, within the field of population genetics, Bartlett notes assumptions are made that aren’t necessarily true in every situation of genetic change. Specifically, fitness is ill-defined, and genetic variation is not always random (scientists have known this for a while).
Modeling Population Dynamics
There are two major ways of modeling genetic variation for population dynamics: origin fixation models and standing genetic variation models. Standing genetic variation models seek to explain current allele frequencies based on pre-existing genetic diversity in the population due to recombination or gene flow. These models largely ignore novel genetic variants.
Origin fixation models assume that novel genetic variants are random events that can contribute to fitness and accumulate in a population. These models aim to describe the process beginning with a novel variant’s introduction, through to its eventual elimination or fixation within the population.
In order to improve modeling of population dynamics, Bartlett argues we must incorporate other known modes of genetic variation. He then makes an attempt at doing this for cyclical genetic variation. He develops a mathematical model for a single locus to model cyclical genetic variation. What he learns from doing this is that the ratio of beneficial to deleterious variants does not depend on the rate at which variants appear, but on the number of genotypes and the sizes of the subpopulations of a specific genotype. This causes an interesting effect.
As long as the cyclical variation rate is not too high, the most reproductively fit individuals dominate the population. As time goes on, most of the variation accumulating in the dominant population is deleterious, moving individuals to less fit genotypes. This means the majority of variants will be leaving this subpopulation and joining a less fit population. So if the environment is constant, then the dominant genotype increases in number and the ratio of beneficial to deleterious variants drops for that subpopulation. In other words, the number of deleterious variants goes up overall in the subpopulation.
However, if the environment changes to a situation where the dominant population is no longer the most reproductively fit, another subpopulation will begin to rise in dominance. That would cause the previously favored subpopulation’s ratio of beneficial to deleterious variants to decrease.
Reframing Genetic Variation
Bartlett has raised excellent points and concerns for the field of population genetics and he has developed an important preliminary model for cyclical mutation at a single locus. The parameterized form of genetic variation discussed in this paper may function as a population-level (think: cloud-based) potential for organismal adaptation, different from random mutation induced by ionizing radiation. Bartlett and other ID proponents hypothesize that conflating this non-random genetic variation with random variation has led to unresolved problems in both population genetics and phylogenetics. Reframing genetic variation as falling into separate categories — random and non-random — could provide new insights into how organisms adapt and how the history of life is interpreted. Future work by Bartlett or others will help shed light on how this can be accomplished.
No comments:
Post a Comment