Science Literature

The premier computer tool for simulating evolutionary processes has the name Avida. Researchers have reported studies on the evolution of complexity, altruism, changes to genetic architecture, and even the evolution of sex. Avida can be used to illustrate numerous concepts and mechanisms of Darwinian evolution and it has achieved a reputation as a platform for carrying out evolution experiments with digital organisms. As someone who has used simulation as a research tool, the question I have always asked (but to which I have struggled to find a satisfactory answer) concerns validation. It appears that Avida authors have made little or no attempt to cross-validate digital evolution with biological evolution. They have not attempted to move out of their virtual reality world by drawing attention to empirical experiments that give credibility to their conclusions. So how does Avida relate to the real world? At last, a paper has appeared which appears to ask similar questions:

" . . . some may ask whether the results of experiments with digital organisms have any relevance to living systems. We conclude that digital genetics is a valid platform for studying some biological questions, but that the applicability of results will depend critically upon the parameters used." (p.11)

The Avida Digital Life Platform demonstrates Darwinian concepts - but does it relate to the real world? (source here)

Inevitably, all simulation models have simplified the system being investigated. This is not, in itself, a problem as long as validation work is undertaken to establish what features in the simulated system can be mapped against the real-world system. So, for example, "key terms such as nucleotide, gene, heritability, selection, and fertility lack a clear equivalent in the software" (p.8). This can be OK, but users must not imagine an equivalence when none exists. Validation work to establish application areas is therefore both relevant and essential.

One problem with Avida has been the high values assigned to beneficial mutation rates and fitness effects.

"Previous experiments using Avida have studied the evolutionary emergence of complex features resulting from high-impact beneficial mutations. Avida's default settings provide mutational fitness effects of 1.0-31.0 for beneficial mutations that give rise to certain computational operations [. . .]. However, fitness effects this large are extremely rare in nature." (p.3)

Avida is constructed according to the Darwinian paradigm. The software is designed so that variations appear using mechanisms of mutation and natural selection. Avida assigns a high proportion of the digital genome to functionless code. Just as Darwinists imagine happens in the real world, so Avida has a genome much of which can be disturbed without disrupting fitness, but which is also capable of experiencing mutations that result in functionality. We now know that this is not a valid representation of the real world. The authors indicate that 85% of the Avida genome is initially benign, but it has the potential to contrinbute to fitness after certain mutations occur. However, as more and more functions are found for junk DNA, this aspect of Avida's design appears increasingly anachronistic.

"Mutations randomly substitute, insert, or delete single instructions in an Avidian genome, drawing upon 26 available instructions defined in the software. The ancestral genome devotes about 15 instructions to the essential replication code, while the remaining 85 positions are occupied by benign no-operation instructions, analogous to inert "junk DNA" that can be used as raw material for evolutionary tinkering." (p.3)

Although the software has been developed with Darwinian mechanisms in mind, the use of more realistic parameters needs attention - principally fitness effects and the proportion of advantageous mutations. The results reported by the authors do not confirm that Darwinian mechanisms can deliver the transformations that Darwinists claim. The problem with gradual incremental evolution is that small advantages are not selected naturally and do not become dominant in the population. Avida assumes and implements a scheme in which complex features can be built step-wise. Whether this is true for biological organisms is another question entirely.

"We observed that, when fitness effects in Avida are small, all advantageous logic operations are lost. Though digital organisms are peculiar in that they can survive such a loss, these data confirm that the accumulation of slightly deleterious mutations can lead to decreasing biological functionality and potentially eventual extinction. Because deleterious mutations are much more common than advantageous mutations in most systems studied, reduction in the efficacy of selection imposes strong directionality on evolution by favoring the fixation of deleterious mutations. The conditions under which fitness recovery may be possible should be studied more thoroughly using computational approaches." (p.12)

The most far-reaching conclusion relates to adaptation. "Plausible" adaptation accounts are pervasive in the literature of Darwinism (see here and here) but few of them are supported by empirical evidence. The Avida findings imply that "plausible" only means "realistic" when mutational fitness effects are large. With only marginal benefits, it is far more likely that "plausible" scenarios turn into "adaptationist just-so-stories".

"In contrast to Avida's default settings, most mutations in biological organisms are low-impact, and this class of mutations may dominate evolutionary change. When Avida is used with more realistic mutational fitness effects, it demonstrates a clear selection threshold. Mutations that influence fitness by approximately 20% or less come to be dominated by random genetic drift. Mutations that affect fitness by 7.5 - 10.0% or less are entirely invisible to selection in this system. These results provide evidence that low-impact mutations can present a substantial barrier to progressive evolution by natural selection. Understanding mutation is of primary importance, as selection depends on the mutational production of new genotypes. Numerous changes that would be beneficial may nevertheless fail to occur because mutation cannot produce them in the time available. Further, it is important for biologists to realistically appraise what selection can and cannot do under various circumstances. Selection may neither be necessary nor sufficient to explain numerous genomic or cellular features of complex organisms." (p.13)

It is worth considering how this research might affect the way evolutionary theory is presented in schools and universities. Adaptation is typically demonstrated by standard examples: peppered moths, Galapagos finches, antibiotic resistance, and more. The logic then appears to be: adaptive change is a real phenomenon so we can proceed on the basis that Darwinian evolution is validated, i.e. it is a plausible explanation for not just some but all features of life. This Avida research allows teachers and students to evaluate this questionable logic and assess the significance of these evidences.
* Are mutations low-, medium- or high-impact?
* What are the evidences relevant to random genetic drift?
* What data are relevant to assessing the relative proportions of low-, medium- and high-impact mutations?
* How important is the concept of "selection threshold" in evaluating potential evolutionary scenarios?
* Does the accumulation of low-impact mutations constitute a health hazard?

At the very least, students can be alerted to the fact that whilst examples of adaptation are documented in the literature, they occur only when mutational fitness effects are large enough. It follows that barriers to adaptive change may be just as important for understanding ecosystems as adaptation. This will take many out of their comfort zone (see here), but they can be encouraged by the thought that they are studying real science, not a virtual reality world.

The effects of low-impact mutations in digital organisms
Chase W. Nelson and John C. Sanford
Theoretical Biology and Medical Modelling, 2011, 8:9 | doi:10.1186/1742-4682-8-9

Abstract:
Background: Avida is a computer program that performs evolution experiments with digital organisms. Previous work has used the program to study the evolutionary origin of complex features, namely logic operations, but has consistently used extremely large mutational fitness effects. The present study uses Avida to better understand the role of low-impact mutations in evolution.
Results: When mutational fitness effects were approximately 0.075 or less, no new logic operations evolved, and those that had previously evolved were lost. When fitness effects were approximately 0.2, only half of the operations evolved, reflecting a threshold for selection breakdown. In contrast, when Avida's default fitness effects were used, all operations routinely evolved to high frequencies and fitness increased by an average of 20 million in only 10,000 generations.
Conclusions: Avidian organisms evolve new logic operations only when mutations producing them are assigned high-impact fitness effects. Furthermore, purifying selection cannot protect operations with low-impact benefits from mutational deterioration. These results suggest that selection breaks down for low-impact mutations below a certain fitness effect, the selection threshold. Experiments using biologically relevant parameter settings show the tendency for increasing genetic load to lead to loss of biological functionality. An understanding of such genetic deterioration is relevant to human disease, and may be applicable to the control of pathogens by use of lethal mutagenesis.

See also:

Anderson, E. Bits, Bytes and Biology: What Evolutionary Algorithms (Don't) Teach Us About Biology, ISCID Paper, November 2, 2004.

Truman, R. Evaluation of neo-Darwinian Theory with Avida Simulations (Part 1 and Part 2), PCID, 3(1), November 2004.