BEACON at GECCO 2016

We are very excited to acknowledge the success of some fellow BEACONites. Kalyanmoy Deb and students of the COIN (Computational Optimization and Innovation) lab this year received multiple awards and nominations at the GECCO conference (The Genetic and Evolutionary Computation Conference).

For the Real-World Applications (RWA) track, Zhichao Lu, Kalyanmoy Deb (Michigan State University) and Ankur Sinha (Aalto University) were nominated for the Best Paper Award for their paper: Finding Reliable Solutions in Bilevel Optimization Problems Under Uncertainties.

In the Evolutionary Multi-objective Optimization (EMO) track, Rayan Hussein and Kalyanmoy Deb (Michigan State University) were nominated for the Best Paper Award for their paper A Generative Kriging Surrogate Model for Constrained and Unconstrained Multi-objective Optimization.

In the Genetic Algorithms track, Kalyanmoy Deb (Michigan State University) and Christie Myburgh (MapTek) won the Best Paper Award voted by the GECCO participants for their paper Breaking the Billion Variable Barrier in Real-World Optimization Using a Customized Evolutionary Algorithm.

20160724_130959 IMG_3625The Niching Methods for Multi-modal Optimization Competition Award was awarded to Ali Ahrari (ME PhD student, Michigan State University), Kalyanmoy Deb (Michigan State University), and Mike Preuss (University of Munster, Germany) for their paper Benchmarking Covariance Matrix Self Adaption Evolution Strategy with Repelling Subpopulations.

Lastly ACM’s SIGEVO Impact Award, recognizing papers previously published 10 years earlier in the GECCO conference that are highly cited and deemed to be seminal by the SIGEVO Executive Committee, was awarded to Kalyanmoy Deb, and J. Sundar for their paper Reference point based multi-objective optimization using evolutionary algorithms in Proceedings of the 8th annual conference on Genetic and evolutionary computation. ACM, 2006.

Congratulations!

Posted in BEACON in the News, BEACON Researchers at Work | Tagged | Comments Off on BEACON at GECCO 2016

BEACON now funds video games?

This is a repost from Terry Soule (Computer Science, UI) and Barrie Robison (Biological Sciences, UI)

Also visit Telliamed Revisited (Richard Lenski’s blog) for another write-up. 

Hello BEACONites,

Thanks to BEACON’s support Polymorphic Games has created the evolutionary video game Darwin’s Demons, and placed it on the Steam website as part of the green light process.

Darwin’s Demons adds an evolutionary component and modern flair to an arcade classic. Darwin’s Demons models biological evolution using enemies with digital genomes. Enemies acquire fitness by being the most aggressive, accurate, and longest lived, and only the most fit enemies pass their genomes to the next generation. The result? The creatures you found hardest to kill have all the babies, making each generation more challenging than the last!

The game includes in-game graphs for tracking evolution, displays the most fit enemies from each wave, and has an experiment mode where you can set parameters like the mutation rate, fitness function, etc. It also dumps all of the evolutionary data to a file. So, there are opportunities for experiments on user driven evolution if anyone is interested. (We are more than happy to share the code and/or make simple modifications for controlled experiments.)

If you get the opportunity please try out the demo (downloadable at either of the sites listed above, with Windows, MAC, and Linux versions), vote for us on Steam, and send us comments, suggestions, or ideas for future directions and collaborations.

Thanks,
Terry Soule (tsoule@cs.uidaho.edu), Computer Science, UI
Barrie Robison (brobison@uidaho.edu), Biological Sciences, UI

Posted in BEACON in the News, BEACON Researchers at Work, Education | Tagged , , , , | Comments Off on BEACON now funds video games?

Culture, Sociality, & Evolution

This Evolution 101 post is by MSU grad student Alex Lalejini

Culture and Chimpanzees

Our species is incredibly social, and one of the major products of our sociality is culture. People typically imagine culture to be exclusive to humans. The idea of culture brings to mind examples that are uniquely human phenomena: blockbuster movies like The Fast and the Furious franchise, great arthouse crime films like Drive, the art exhibited at the Louvre in Paris, or sharing a traditional meal with one’s family. These phenomena are perfectly valid associations with the idea of culture, but it may come as a surprise that culture is not exclusive to our species.

Dr. Andrew Whiten, a professor at the University of St. Andrews and an expert on non­human cultures (yes, non-­human!), has done extensive work on chimpanzee culture. Chimpanzee culture! But we don’t see chimpanzees publishing novels or displaying art in galleries. This observation is likely correct; however, one must consider the definition of culture. Whiten defines culture as “all we learn from others that endures for long enough to create socially transmitted traditions” (Whiten, Vince, & Mace, 2014). With this definition, cultural behavior is well established in species other than humans: dialects in song­-birds, sweet-­potato washing and stone-­handling by Japanese Macaques, and tool use and various other social behaviors in African chimpanzees (Whiten et. al., 1999). By the way, you can hear about some of Whiten’s work on chimpanzee culture in a presentation here.

Culture is a deeply intriguing social phenomenon. How and why does cultural behavior emerge from evolutionary processes, and what advantages does cultural behavior bestow upon its practitioners? The journey to understanding the evolution of cultural behavior meanders through the evolution of sociality. Basic forms of social interactions form the building blocks for cultural behavior. Without sociality, individuals lack the capacity to socially transmit information among one another, and without the capacity to socially transmit information, cultural behavior cannot exist. So, what exactly does it mean to be social? What other forms of social behaviors are found in nature? And, why might social behavior evolve in the first place?

What does it mean to be social?

For the purpose of this discussion, we are going to consider sociality as a tactic employed by an individual or group of individuals that increases the fitness of group members (Armitage, 1999). Evolution has produced many examples of sociality in organisms; aside from cultural behavior, this blog post will give an overview of two: social grouping and parental investment. Social grouping can be seen in herds of African elephants, schools of surgeonfish, or flocks of seagulls. Parental investment is defined as “any investment by the parent in an individual offspring that increases the offspring’s chance of surviving (and hence reproductive success) at the cost of the parent’s ability to invest in other offspring” (Campbell, 1972). Parental investment can be seen in mother grizzly bears taking care of their cubs. Rather than abandoning her offspring at birth in favor of having more offspring, the mother grizzly takes care of her cubs until they reach sufficient maturity to take care of themselves.

Why be Social?

All of these examples are great, but if we think back to natural selection, an individual’s fitness is a function of their ability to reproduce relative to their competitors. If one individual is acting socially to increase the reproductive success of other group members, doesn’t sociality decrease an individual’s fitness relative to other group members, and shouldn’t natural selection oppose social behavior? At a glance, this seems like a reasonable hypothesis; however, with a closer look, one can see the evolutionary benefits of organismal sociality.

Social grouping can decrease an individual’s susceptibility to predation through increased vigilance (example: shared responsibility for watching for predators in herds), aggressive group defense behavior, increased ability to find and obtain food sources, or just through the fact that larger groups lower an individual member’s probability of being preyed upon (Swedell, 2012; Alexander, 1974).

Parental investment is a deeply interesting area of research, and is much more complicated in actuality than what I will mention here (see the parental investment section of Alexander 1974 for a nice introductory overview). Kin selection offers an insightful view into some of the advantages of parental investment. Kin selection is “the evolution of characteristics which favour the survival of close relatives” (Smith, 1964), which is potentially at the cost of the individual. In other words, if an individual’s close relatives are reproductively successful with the aid of the individual – even at the cost of the individual – the individual’s genes are likely to persist into the next generation because the individual and their close relatives are likely to share many genes. In the context of parental investment, parents may invest heavily in their offspring in order to increase the possibility of grand-­offspring.

We can clearly see how social behaviors like social grouping and parental investment facilitate cultural behaviors by looking to our own species. The social transmission of traditions from parent to offspring and from individual to social group is widespread in human societies; this transmission can be thought of as an instance of parental investment. We also often share socially acquired knowledge and traditions with other people in our local communities (e.g. your grandmother’s famous recipe for gumbo); these communities act as our social groups and provide a forum to socially transmit information to others. Social grouping and parental investment might be potential building blocks for the evolution of cultural behavior, but what possible selective advantages are conferred to practitioners of cultural behavior?

Why be Cultural?

If we consider culture and organisms’ genomes as information storage mechanisms, culture provides an alternative method of encoding and passing information among individuals in a population as compared to gene propagation through reproduction. As an information storage mechanism, culture is much more fluid and quick to change than information stored at the gene level (there is no need to wait around for mutation and natural selection to act); however, this is at the cost of being more fragile (traditions must be taught to offspring, there is no way to pass them down genetically).

Insights into the evolution of sociality may lead to a deeper understanding of evolutionary history because many species behave socially. Our own species is incredibly social; you can imagine how the human species displays all of the examples I’ve talked about here (cultural behavior, parental investment, social grouping). Progress in our understanding of the evolution of sociality may offer us an even deeper understanding of our own species.

References

Alexander, R. D. (1974). The evolution of social behavior. Annual review of ecology and systematics, 325­383.

Armitage, K. B. (1999). Evolution of sociality in marmots. Journal of Mammalogy, 80(1), 1­10.

Campbell, B. G. (Ed.). (1972). Sexual selection and the descent of man, 1871­1971. Heinemann.

Smith, J. M. (1964). Group selection and kin selection. Nature, 201, 1145­1147.

Swedell, L. (2012). Primate sociality and social systems. Nature Education Knowledge, 3(10), 84.

Whiten, A., Goodall, J., McGrew, W. C., Nishida, T., Reynolds, V., Sugiyama, Y., … & Boesch, C. (1999). Cultures in Chimpanzees. Nature, 399(6737), 682­685.

Whiten, A., Vince, G., Mace, R, (2014). Human and Other Animals: Cultural Evolution and Social Learning. Lecture given at The Royal Institute

Posted in Evolution 101 | Tagged , , | Comments Off on Culture, Sociality, & Evolution

Genetic Drift

This Evolution 101 post is by MSU grad student Ali Tehrani

Darwinian evolution suggests that the pattern of changes in populations is driven by natural selection, i.e. those individuals that are more fit survive and reproduce. Does that mean populations of organisms are constantly gaining beneficial genes and losing detrimental genes through natural selection? So, is it not possible that some genes that are actually beneficial or at least neutral to the organism’s fitness have disappeared in evolutionary history only by chance? Or alternatively that random events could cause slightly deleterious genes to increase in frequency?

There are several examples in nature in which beneficial or neutral traits have disappeared from a population. Consider an island with a population of ten rabbits. This island can only hold a small population of rabbits (about 10) due to resource limits. Five rabbits out of ten are white and the other five are brown (figure 2). Also assume that the coat color does not have any effect on their reproductive success. Coat color is a genetically inherited trait with two different alleles1, one for white coat color and the other for brown. If both parents have the same coat color (i.e. both of them are white or both are brown), their offspring will also have that coat color (same as their parents). If one of the parents is white and the other is brown, their offspring will be either white or brown with equal probabilities (see figure 1). Rabbits in this population mate randomly; thus, the probability of mating two white rabbits is the same as the probability of mating between two brown rabbits. So, there is a fifty percent chance that each individual in the next generation will be white or brown. Does this mean that the population of rabbits will always have five white rabbits and five brown ones?

Figure 1. Coat color in offspring of rabbit couples. If both parent 1 and parent 2 are white, all their offspring will be white (top left). If both parents are brown, all their offspring will be brown (bottom right). If one parent is white and the other is brown there is a 50% chance that their offspring is white or brown (bottom left and top right).

In reality, there is no guarantee that such system will always have five rabbits of each coat color. It is similar to flipping a coin ten times and getting different numbers from five heads and five tails. This phenomenon, which is called “sampling error”, results in fluctuations in number of rabbits with white/brown coat color. Sampling error is more likely to happen in small populations because in these populations, obviously, the sample size is smaller. The random fluctuations in allele frequency in a population which occurs due to sampling error is called “genetic drift”. The effect of genetic drift is stronger in small populations. Sampling error becomes smaller as the population size grows for the same reason that if we flip a coin one million times the deviation from 50%-50% becomes much smaller.

Figure 2. Initial population of rabbits consisting of five white-colored and five brown-colored rabbits.

Let’s go back to our rabbit population. Given the initial population in figure 2, and the mating pattern illustrated in figure 1, each individual in the second generation would be brown or white with equal probability of 50%. So how many white rabbits would be in the second generation? Well, this is similar to ask: how many heads do we get if we flip a coin ten times? In reality, it is possible that we get any number of heads. So assume we flipped the coin ten times and we got six heads and four tails.

In a similar manner, suppose we have six white rabbits and four brown rabbits in the second generation (figure 3). Now, the probability of a particular rabbit being white in next generation is not 50% anymore (it is higher than 50%). This deviation can become either larger or smaller each generation, randomly, and it is possible that eventually the population becomes uniformly white- colored (figure 3). When all individuals in the population have the white coat color allele, this allele is said to be “fixed” in the population. Because there is no rabbit with a brown coat color allele in the population anymore, it is therefore impossible for this allele to appear in the population again in the absence of mutation and migration. As a result, the frequency of the white coat color allele becomes 100% and remains the same thereafter (figure 4 shows how coat color allele frequency changes through generations).

Figure 3. Fluctuations in the number of white/brown rabbits through generations. Genetic drift in this scenario led to fixation of white coat color allele in the population.

Genetic drift and consequent random fixation of alleles in populations can occur as a result of abrupt reduction in population size. The abrupt decline in population size which may occur due to environmental effects (e.g. drought, fire, flood, etc.) is called “population bottleneck”. Genetic drift is a significant evolutionary process in such small populations and usually leads to fixation of alleles to one or the other. Hence, the genetic diversity within such populations decreases or is totally lost as a result of genetic drift operating on small number of individuals founding the population.

Figure 4. Coat color frequency of rabbits through generations.

[1] https://en.wikipedia.org/wiki/Genetic_drift
[2] Conner, Jeffrey K., and Daniel L. Hartl. A primer of ecological genetics. Sinauer Associates Incorporated, 2004.
[3] https://en.wikipedia.org/wiki/Population_bottleneck

1 An allele is a variant form of a gene. In this example, the gene for white coat color and the gene for brown coat color are two alleles of the coat color gene. For the purposes of this example, we will ignore the fact that rabbits are diploid — containing two alleles of every gene, one inherited from each parent — to simplify the math.

Posted in Evolution 101 | Tagged , | Comments Off on Genetic Drift

Evolution of Reliable Signals

This Evolution 101 post is by MSU grad student Thassyo Pinto

The ownership of goods such as luxury cars, expensive boats and conspicuous consumption, and showing it off to others, transmits a signal informing that owner is capable of bearing expenses. The wealth and status are used to advertise “fitness”, which demonstrates higher quality to potential mates. In the Martin Scorsese’s film “The Wolf of Wall Street”, Leonardo DiCaprio (Figure 1) takes the role of a crooked stock trader who has a great talent for conspicuous consumption. As said by his character, Jordan Belfort: “Their money was better off in my pocket – I knew how to spend it better”. Not only did his character spend a fortune on extravagant goods and illegal drugs, but he also used his status to mate with multiple partners. Since the cost of conspicuous consumption is high, in theory there should be few cheaters (underprivileged men) who attempt to mimic these “honest” signallers in order to attract women.

Figure 1. A scene from the movie “The Wolf of Wall Street”

In the animal kingdom, signals are important characters since them convey various information from signallers to receivers. The reliability of a signal can change depending on the receiver responding to reliable signals and ignoring non-reliable ones. Moreover, the cost of a signal could affect its reliability since the investment in it could be worthwhile to honest signallers and not worthwhile to cheaters. We can take a peacock’s tail for example (Figure 2), which provides positive value to the female. If this signal is really costly (a long and heavy tail), a male not only could look attractive to a female, but the signal could also tell that he has a better chance for survival. As long as a certain quality is not impacting a male’s health, he must be well-adapted in comparison to less adorned ones. This idea was first proposed by Amotz Zahavi (Zahavi, 1975) and it is called the handicap principle.

Figure 2. Male Indian peacock (Pavo cristatus), displaying colorful and vibrant feathers

In Zahavi’s handicap principle, the handicap not only is an indicator of genetic quality, but it also needs to be costly in order to ensure that signalling is reliable. Otherwise, low quality males would be able to equally advertise and females would not be able to identify the honest signaller. Although this original idea was largely rejected by many scientists in its earlier days, Alan Grafen’s (Grafen, 1990) paper provided the first full game-theoretical formalization, confirming the plausibility of the handicap principle mathematically. Figure 3, shows how tail length relates to costs and benefits. A costly tail (very long) will be more affordable to high quality (HQ) males than low quality males (LQ). If there is a benefit in tail length due to mating success, the optimal tail length will be higher for a high quality males than for low quality males. Furthermore, if low quality males try to grow a much longer tail than its optimal value, his benefit will decrease, showing that the tail length becomes an honest signal. Therefore, the cost creates a correlation between male quality and tails cost, making a cheating behavior maladaptive.

Figure 3. Optimal tail length for a low quality male (LQ) and for a high quality male (HQ) when a long tail is a handicap (Lotem, 1993).

Signals can also be directed from prey to predator. Stotting gazelles send signals to predators conveying that they have a lot of energy and they are in good health condition. As described by Zahavi (Zahavi, 1997), “it shows off its strength and fitness by jumping straight up. Only a gazelle certain of its ability to outrun a predator dares squander its strength is this way”. These hard-to-catch prey benefit by differentiating themselves from the overall population, and consequently deter predation. If there was no cost in this signal, there would be a growing number of cheaters, and predators would learn to ignore these false signals.

As suggested by Zahavi (Zahavi, 2008) “the signal encodes neither threat nor invitation, but rather dimensions of a quality, i.e. strength, which produces different reactions in different receivers”. Zahavi also infer that signalling systems are by nature collaborations. In order to a signal to be effective, there must be a cooperation between the receiver and the signaller. Therefore, Zahavi implies that the handicap principle guarantee the reliability of signals and it is an essential component in all signals.

References:
Grafen, Alan. “Biological signals as handicaps.” Journal of theoretical biology 144.4 (1990): 517-546.

Lotem, A. “Secondary sexual ornaments as signals: the handicap approach and three potential problems.” Etologia 3.209-18 (1993).

Zahavi, Amotz. “Mate selection—a selection for a handicap.” Journal of theoretical Biology 53.1 (1975): 205-214.

Zahavi, Amotz, and Avishag Zahavi. The handicap principle: a missing piece of Darwin’s puzzle. Oxford University Press, 1997.

Zahavi, Amotz. “The handicap principle and signalling in collaborative systems.” Sociobiology of communication. Oxford University Press, Oxford (2008): 1-11.

Posted in Evolution 101 | Tagged , , , , | Comments Off on Evolution of Reliable Signals

Artificial selection and correlated traits

This Evolution 101 post is by MSU grad student Tyler Derr

One of the basic mechanisms of evolutionary change is natural selection. It was in Charles Darwin’s famous book, On the Origin of Species, where he defined natural selection to be the “principle by which each slight variation [of a trait], if useful, is preserved.” (Darwin 1859). Knowing that many people would be skeptical of what he presented in his book, the first chapter is structured to first discuss selection in terms of breeding. He presented examples of where humans have selectively bred both animals and plants.

Artificial selection (also called selective breeding) is a process in which humans interfere with natural selection to obtain certain traits we desire an organism to have. This process is performed by choosing which animals or plants are allowed to mate with each other in the hopes if both of the parents have a certain observable trait then their offspring will as well. Shown in Fig. 1 are common vegetables that have all been cultivated from wild mustard by past farmers artificially selecting traits in the plant.

Fig. 1. Vegetables that arose from the wild mustard due to selecting for different traits.

Selective breeding has been done on many animals. Examples of such artificial selection being performed can be seen in dogs. Due to this selective breeding, there are now hundreds of different breeds. It was only just recently, thanks to whole genome sequencing, that we have discovered that gray wolves and dogs started to diverge from a common ancestor at the same time roughly 27,000- 40,000 years ago (Skoglund 2015). Although today we breed dogs for certain traits such as cute floppy ears or a fluffy coat, we can pretty safely say that these would have probably not been at the top of the list with humans that long ago. We know that based on when dogs first arose this would have been the time when humans were still hunter-gathers (Freedman 2014). Fig. 2 below shows a few dog breeds representing were they were selectively breed, based on geographical location, and also representing from what previous dog breeds they were breed from.

So if humans tens of thousands of years ago did not select for all the same traits that we select for today, what did they desire? Well, it can be hypothesized that they simply desired the dogs to be non-aggressive. Although we normally think of evolution as a very long process (which it is with natural selection), the game changes when we as humans start getting involved. It was in 1959 that a Russian geneticist Kmitry K. Belyaev (shown in Fig. 3) began a study in the hopes of breeding a population of tame foxes (Trut 1999). Belyaev solely selected for tameness and strictly against aggression when breeding. After just ten generations, 18% of the pups were not only tame, but showing signs of affection such as whimpering for attention and even licking the experimenters (Trut 1999). Evolution in Action: The Silver Fox Experiment, is a short clip from the BBC’s documentary [The Secret Life of Dogs] in which they discuss the early stages of the fox experiment and then also show the progress that has been made over the last 50 years.

Fig. 2: Examples of how selective dog breeding has branched from the common ancestor based on geographical location. Cain, Michael L., Damman, Hans, Lue, Robert A. and Carol Kaesuk Yoon. Discover Biology Second Edition. New York: W. W. Norton, 2002.

As mentioned before, the only trait that was selected for in the fox experiment was tameness, however interestingly enough quite a number of other traits came with it. Other physical traits such as a curly tail instead of straight, floppy ears, and shorter limbs began to appear in the tame foxes, which are traits commonly shared among other domesticated animals (Trut 1999). What was just described is commonly known as the correlation of traits, which is another interesting topic to mention when discussing artificial selection. This is when selecting for a specific trait not only allows the offspring to have the trait selected for, but also a set of other traits that are genetically correlated.

Fig. 3: Belyaev shown with some of the tame foxes that were bred in his experiment.

 

As seen from both the plant and animal examples, humans can have a large impact on a species when performing selective breeding. It will be interesting to see how future organisms will be changed to fit the wants and or needs of humans in the future.

References:

Darwin, C. (1859). “On the origin of species by means of natural selection, or by preservation of favoured races in the struggle for life”. London: John Murray.

Freedman, A., et al. (2014). “Genome sequencing highlights the dynamic early history of dogs”. PLoS genetics 10(1): e1004016.

Skoglund, P., et al. (2015). “Ancient wolf genome reveals an early divergence of domestic dog ancestors and admixture into high-latitude breeds”. Current Biology 25(11): 1515–9.

Trut, L. (1999). “Early Canid Domestication: The Farm-Fox Experiment Foxes bred for tamability in a 40-year experiment exhibit remarkable transformations that suggest an interplay between behavioral genetics and development”. American Scientist 87(2): 160-169.

 

Posted in Evolution 101, Uncategorized | Tagged , | Comments Off on Artificial selection and correlated traits

It’s a (Selective) Sweep for the Good Genes!

This Evolution 101 post is by MSU grad student Douglas Kirkpatrick

Fig. 1: Lebron James helps lead the Cavaliers in a sweep of the Celtics.

In baseball, ice hockey, and basketball, when a team wins all of the games in a playoff series, they are said to have swept the other team out of the playoffs. In other words, a sweep is the complete victory of one group over the other, often due to a star player. Funnily enough, this terminology applies almost as well in the world of evolutionary biology. A selective sweep is the process by which strong selection of a beneficial allele causes the surrounding linked alleles to achieve a higher frequency in the population. In the same way that superstar Lebron James lead his team, the Cleveland Cavaliers, to a sweep of the Hawks and the Celtics in last year’s playoffs, a good gene might help linked alleles attain reproductive success in a selective sweep.

The essence of this process is that when a hugely beneficial allele appears, it is rapidly selected for. Due to the rapid selection, many linked or nearby genes are also passed on to descendants, even though these other alleles may be neutral or even mildly deleterious to the organism.   This process is illustrated in Figure 2. The extremely beneficial gene is highlighted in red, while deleterious or neutral genes are presented in blue. Each line in Figure 2 represents the genome of a single member of the population.   The pattern of the beneficial gene and its neighbors, seen in the fourth row on the right, are collectively referred to as a “haplotype.” The haplotype of the advantageous gene becomes much more frequent after selection occurs, as new populations are spawned [3].

Fig 2: A basic biological selective sweep

Several methods exist to detect selective sweeps; the primary method is to measure linkage disequilibrium. That is, the distributions of alleles in a population are compared to determine the presence of haplotypes. If a specific haplotype, or collection of linked alleles, is exceedingly common, then it is likely that a selective sweep occurred in the recent past. The linked alleles are primarily those that are collocated within the genome. An alternative measure to find instances of selective sweep is to measure the time to most recent common ancestor. If they have all evolved from a recent ancestor, then it is likely that a selective sweep has occurred. That is, because the ancestor is recent, its haplotypes have likely spread quickly through the population, making it probable that a selective sweep has occurred.

Two examples of selective sweeps are particularly relevant in the modern world: in pathogenic bacteria and in agriculture. Bacteria and disease-causing bacteria in particular, have a short life span. As such, any new allele that creates a more virulent form will spread wildly as the less potent haplotypes die off in a short time. The potential for rapid change in addition to strong selective pressure from outside forces like antibiotics or antivirals leads to many selective sweeps. The short life cycles and high selective pressures placed on pathogens have been shown to have caused selective sweeps in influenza and toxoplasma gondii [1].

While the selective sweeps in bacteria may be detrimental to mankind, there are selective sweeps that have helped humanity. Data shows that selective sweeps were responsible for unifying a diverse population into what we know as modern corn. Artificial pressure, driven by farmers only choosing optimal offspring, in coordination with selective crossbreeding forced a fast evolution of corn. The optimal traits selected for and the strong evolutionary pressure created a selective sweep [5]. The selective sweep in this case proved to be largely beneficial for people.

Figure 3: Selective Sweeps in human populations. Tishkoff et. al, 2007

Numerous examples of selective sweeps can be found in human DNA. There are at least six different chromosomes that show evidence of selective sweeps [4]. The most noticeable was described by Tishkoff et al., and comes as a result of the alleles for lactose tolerance. There was strong selection for the alleles that allow humans to digest milk as adults. As a result, selective sweeps occurred; this is illustrated in Figure 3. Interestingly, similar mutations occurred separately in Africa (Group A) and Eurasia (Group B). Each individual bar represents a portion of an individual’s genome that is shared with the group; a larger bar indicates that more genetic material is the same as the original cell. The horizontal axis is the relative position to the allele for lactase tolerance. The red and the green in each graph respectively show the haplotype for tolerance that has become more common, while the blue and orange are an old haplotype that has been outcompeted [2]. This information is essential to understanding how different yet similar processes evolved in humans; it is likely to be important going forward due to how many selective sweeps have occurred in the human genome.

Selective sweeps are important because they allow for rapid evolution in a short period of time. The strong selection for a specific haplotype can quickly change the distribution of alleles in a population. In addition, finding a selective sweep can help identify key periods of evolutionary change. Sweeps have had a major impact both on the human genome, and that of plants and animals seen in everyday life; they will continue to do so into the foreseeable future.

Sources:

  1. Sa, Juliana Marth, et al. (2009). “Geographic patterns of Plasmodium falciparum drug resistance distinguished by differential responses to amodiaquine and chloroquine”. PNAS 106 (45): 18883–18889.
  2. Tishkoff, Sarah A., et al. “Convergent adaptation of human lactase persistence in Africa and Europe.”Nature genetics 1 (2007): 31-40.
  3. A selective sweep. http://www.nature.com/scitable/content/a-selective-sweep-24827
  4. “A haplotype map of the human genome”. Nature 437 (7063): 1299–1320. October 2005. doi:10.1038/nature04226
  5. Gore, Michael A., et al. (2009). “A First-Generation Haplotype Map of Maize”. Science 326: 1115–1117. doi:10.1126/science.1177837. PMID 19965431.
Posted in Evolution 101 | Tagged , , , , , | Leave a comment

Pseudogenes

This Evolution 101 post is by MSU grad student Tyler Derr

I’m sure you’ve heard the saying that our DNA is the “blueprint” of who we are. Well, our genes are the sequences in our DNA that actually encode instructions for particular functions. What might come as a surprise to some is that in humans over 98% of our DNA is non-coding regions (Elgar and Vavouri 2008). This means that less than 2% of our DNA actually codes for proteins. Pseudogenes, unlike our genes, fit into the non-coding category, but what exactly are they?

A pseudogene is defined as a sequence in our DNA that is homologous to a known gene, but is nonfunctional (i.e. looks like a gene, but for some reason or another, just can’t quite make the cut to creating a functional protein). Therefore it would seem logical to assume, since they can’t make a functional protein, that they serve no purpose. In fact, until just a few years ago, scientists believed that they didn’t have any immediate function. However, now we know that some pseudogenes can actually serve an important function to an organism and not just a critical role in evolution. After discussing the types of pseudogenes and an example of how they can sometimes provide an immediate useful functionality to an organism, we will discuss their relevance to evolution.

There are three main types of duplicated pseudogenes: unitary, duplicated and processed and duplicated non-processed (Figure 1). This categorization is based on how the pseudogenes appear. We shall further discuss these three types below.

Figure 1: Types of duplicated pseudogenes: unprocessed (top) and processed (bottom). Note that they are all created in some way from a functional gene. Rouchka, Eric C., and I. Elizabeth Cha. “Current trends in pseudogene detection and characterization.” Current Bioinformatics 4.2 (2009): 112-119.

A processed duplicated pseudogene occurs during what is called retrotransposition. This is when a portion of mature mRNA is placed back into the DNA. This type of pseudogene is the easiest to detect in our DNA due to the fact that this process (also known as reverse transcription) allows the insertion of the mRNA poly-A tail into the DNA. The poly-A tail is just a long sequence of ‘A’s, which is a characteristic of mature mRNA and is usually not found in our DNA. One of the reasons why these insertions are classified as pseudogenes is because the placement of the mRNA in the DNA lacks a promoter sequence, which acts as a flag that represents where to start the transcription process. The steps required for making a processed pseudogene can be seen in the bottom section of Figure 1.

The second type of pseudogenes are unprocessed duplicated pseudogenes and are created during the copying of genes in the DNA (gene duplication). Once a gene has been duplicated, if one of the gene copies incurs a mutation, such as a nucleotide change that results in an early stop codon in the middle of the gene, it loses its ability to code for a protein and can be thought of as “junk DNA”. Normally, there would be a huge selection pressure on such a mutation if a single gene no longer functioned, but since the mutation happened on a gene that had undergone gene duplication, there would still be at least one functioning copy of that gene in the genome. Thus, this type of pseudogene can undergo genetic drift and acquire more mutations that have no direct effect on the fitness of the individuals that have this DNA sequence. The duplication and mutation steps are shown in the upper section of Figure 1.

The last type are the same as the duplicated pseudogenes in that they occur due to mutations, but instead of happening to a gene that has undergone gene duplication, unitary pseudogenes are when the mutated gene is the only copy of itself in the genome. The argument used as to why there would not be selection pressure on individuals that have a duplicated pseudogene no longer holds with the unitary pseudogenes. This is because if the mutated gene has no duplicates, then the gene has been completely deactivated.

As mentioned earlier, scientists used to consider pseudogenes in the category of “junk DNA”, nonfunctional gene lookalikes, and knew them mostly as just sequences that caused problems in their studies (e.g. PCR experiments). However, over time and with further study, quite a number of surprisingly interesting findings have been uncovered involving pseudogenes. One specific example from 2010 that was published in Nature discovered that what we had previously known as a pseudogene, PTENP1, was in fact helping suppress tumor growth in many colon cancer cell lines (Poliseno et al. 2010). The basic idea is that although PTENP1 could not undergo translation to become a functional protein, it was able to play the role of a decoy in having microRNA bind with its processed mRNA rather than the mRNA processed from the PTEN gene. This allowed for more PTEN protein since if the microRNA had attached to the PTEN mRNA it would have not been able to undergo translation to becoming a protein. This process is shown below in Figure 2. The authors had proposed that PTENP1 be no longer considered a pseudogene, but instead a “bona fide tumor suppressor gene.”

Figure 2: Visual explanation of how the pseudogene PTEN1 can help suppress tumor growth. Image Credit: hms.harvard.edu

Even though we just discussed an example where a pseudogene performs an immediate function for an organism, we also need to mention how they can function beyond the scope of an individual organisms lifetime, and instead on an evolutionary scale. In fact (as mentioned earlier) a duplicated pseudogene has the potential to undergo genetic drift and acquire multiple mutations with no detrimental effects to the fitness of the individual. It can be the case that this sequence of DNA later be resurrected into a gene resulting in a new functional protein being formed (Zhang 2003). It can sometimes be the case that simple (e.g. single nucleotide) mutations can result in huge jumps in the functional space; even to the point where the gene is turned off. It might come as a surprise, but theoretically having multiple mutations (which in comparison results in a larger step in the DNA sequence space) on a duplicated pseudogene can actually result in a smaller step in the functional space. These multiple mutations can result in a fitness improvement to an individual. Thus showing that pseudogenes provide an avenue for genetic drift to take place and due to the resurrection of the mutated duplicated gene an individual can express the a function that provides a fitness improvement.

We have discussed the three main types of pseudogenes, introduced an example that has proven some pseudogenes are actually important in their current state, and discussed reasons as to why they are valuable from an evolutionary standpoint. Although it might not be the case that every pseudogene has a current unique and important function to fulfill, those that currently do not have a purpose are still undergoing genetic drift and could possibly, one day, arise to serve a purpose for our ancestors many years from now.

References:

Elgar, G. and Vavouri, T. (2008). “Tuning in to the signals: noncoding sequence conservation in vertebrate genomes”. Trends in genetics, 24(7), 344-352.

Poliseno, L., et al. (2010). “A coding-independent function of gene and pseudogene mRNAs regulates tumour biology”. Nature, 465(7301), 1033-1038.

Zhang, J. (2003). “Evolution by gene duplication: an update”. Trends in Ecology and Evolution, 18(6), 292-298.

 

Posted in Evolution 101 | Tagged , , | Leave a comment

Selfish Genes and the Resulting Gene Conflict

This Evolution 101 post is by MSU grad student Alex Lalejini

The above comic strip might lead one to believe that the phrase ‘selfish genes’ describes genes that make individuals act selfishly; however, this is not at all what is meant by the phrase ‘selfish genes’. This post gives a brief introduction to selfish genes, which is a story rich in greed, replication, and conflict. Aside from the greed, replication, and conflict, the effects and implications of selfish genes are far-reaching, which makes them incredibly interesting.

The Gene Perspective

We are accustomed to thinking about evolution from the perspective of whole organisms: Individual organisms in a population have varying observable characteristics, or phenotypes, as a result of inherited genetic variations. Some variations increase an individual’s ability to compete for resources and reproduce, and through natural selection over many generations, the beneficial genetic variations are propagated through the population of organisms.

However, it is interesting to consider evolution from a more gene­centered perspective. Richard Dawkins has described living organisms as “throwaway survival machines for genes” with this gene­centered perspective in mind. While individual organisms inevitably die, the information coded by their genes has the chance to continue from generation to generation. Genes with phenotypic effects beneficial to a host organism’s chance of survival and reproduction have an improved chance to persist over many generations as compared to genes with less beneficial or harmful phenotypic effects. This is not always the case; some genetic elements achieve persistence and propagation in a population without any consideration for their host organism.

Selfishness

Genetic elements may exploit alternative methods of persistence and propagation without contributing to their host organism’s fitness. As such, these genetic elements are not necessarily invested in the host organism’s fitness, and as a result, their alternative methods of propagation often negatively affect the host organism’s fitness. These genes are selfish – their expression advances their own interests at the expense of other genes and the host organism as a whole. There are a multitude of types of gene selfishness. Here I provide an overview of one: transposable elements.

From Genes in Conflict by Burt and Trivers (2009), transposable elements “accumulate by copying themselves into new locations of the genome”; they are often referred to as selfish DNA parasites. Transposable elements are analogous to a self­replicating computer virus that copies itself many times when it is put on a computer in order to avoid removal. As we are aware, this type of computer virus could have a significant negative effect on the performance of a computer. In a similar manner, transposable elements can have a strongly negative effect on the host organism’s fitness. This form of selfish expression is surprisingly widespread – at least 45% of the human genome is derived from transposable elements (Lander et al.,2001)!

On a bit of a historical note, we must thank Barbara McClintock and her work with the familiar Thanksgiving holiday staple, multi­colored corn, for the discovery of transposable elements. As McClintock discovered, transposable elements are responsible for the vivid mosaics of color seen in Indian Corn. For those with further interest in transposable elements, her work is a great place to start.

Conflict

We have now seen that the expression of selfish genes can increase their own fitness at the cost of other genes and the host organism. As a result, selfish genes are often at odds with the ‘unselfish’ genes that rely on the reproductive success of the host organism in order to increase in frequency. This tension caused by opposing interests facilitates a form of gene conflict. If there is a selfish gene negatively affecting organismal fitness present in a population, there is selection pressure for other genes that suppress the selfish genetic element’s expression. As a result, we see genes with contradictory or conflicting effects evolve.

Evolutionary Implications

While the immediate effects of selfish genetic elements on host organisms are often negative, there is some evidence to suggest that selfish genetic elements and the resulting gene conflict helps to drive evolutionary change and innovation (Werren, 2001). As some genetic elements evolve to get ahead at the expense of the rest of the organism, other genetic elements arise to minimize the negative effects of selfish genes. In the case of transposable elements, the rest of the genome is sometimes able to recruit a transposable element for new cellular functions (Werren, 2001). In this way, selfish genetic elements can be instrumental in pushing organisms toward increasing genetic robustness. Selfish genetic elements, however, do not always bestow long­term positive effects on organisms; they can also lead to species extinction, which perhaps ironically, also leads to the selfish genetic element’s extinction.

References

Burt, A. & Trivers, R. (2009). Genes in conflict: the biology of selfish genetic elements. Harvard University Press.

Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C., Baldwin, J., … & Grafham, D. (2001). Initial sequencing and analysis of the human genome.Nature,409(6822), 860­921.

Werren, J. H. (2011). Selfish genetic elements, genetic conflict, and evolutionary innovation. Proceedings of the National Academy of Sciences,108(Supplement 2), 10863­10870.

Posted in Uncategorized | Leave a comment

Evolution 101 – Mutations: From the X-Men to the X-Chromosome

This Evolution 101 post is by MSU grad student Douglas Kirkpatrick

Fig 1: The X-Men

Everyone knows what mutation is, right? It’s that magical scientific hand-wave that gives the X-Men their powers. Almost certainly the result of interaction with gamma radiation or toxic substances, mutation always has the most drastic results on the people that it affects. Right?

Unfortunately for us would-be superheroes, these results are neither likely nor commonplace, and the effects of mutation are almost always exaggerated while its everyday occurrence is typically ignored. The best example of how mutation affects our everyday life is, ironically enough, found in a brief scene from X-Men: First Class. The young Professor Xavier mentions while wooing a young woman that based on her blue eyes alone she is a mutant, as blue eyes are a result of a specific mutation to DNA. This observation is as close to science fact that the movie comes.

The word mutation comes from the Latin “mutare” which means “to change.” [2] Thus, by its roots the word mutation indicates a change or modification to something. In the context of biology, this is specifically a change to the genetic structure of an organism, to the organism’s DNA. Within that broad definition fall many forms of mutations, each with varying results and effects. An overarching classification for mutations is the scale at which they operate. Small scale mutations affect only a single gene, or perhaps only a few DNA base pairs, whereas large scale mutations affect an entire chromosome or chromosomes.

The small scale mutations fall into one of three main types. First are the point mutations, where a specific DNA base is changed to a different base, from Adenine to Cytosine perhaps. These are potentially caused by radiation, but are often prevented due to the matched base pairs of DNA and the assistance of corrective proteins. The second small scale mutation is addition, where a sequence of DNA is added to a given location in the chromosome. The third, deletion, is the opposite of addition. Deletion occurs when a certain sequence of DNA is removed from a given location. Deletion is visually demonstrated in figure 2. Despite their size, these mutations can still have a large effect on the organism. A single point mutation in hemoglobin, causing the replacement of an amino acid, is the source of sickle cell disease. [1]

Large scale mutations can be similarly subdivided. Duplications and deletions occur when regions of the genome are repeated or removed, respectively, from a chromosome. These types of mutations can cause a certain protein to be overproduced or not produced at all. Underproduction of a protein might result in toxic waste buildup that the cell cannot clean up, while overproduction might result in the cell transporting away all of its raw materials; neither is beneficial to cell behavior. A visualization of these mutations can be found in figure 2. Translocation occurs when portions of the genome move from one location to another.   This can happen between different locations on the same chromosome, or when a gene moves from one chromosome to an entirely new chromosome. The demonstration of movement to another chromosome can be seen in figure 2. Inversion results when direction of a region of the genome is reversed, making it effectively unusable. Just as long passages of text are difficult to read backwards, inverted genes can no longer be read and used to produce proteins. On the extreme end of this classification is the addition or removal of entire chromosomes, known as aneuploidy. A good example of this is the replication of the 21st chromosome that leads to Down Syndrome.

Further classification of mutations is possible, such as the effect on the organism, or alternatively the effect on cellular function. Effects on an organism can be beneficial, and aid the fitness of the organism; detrimental, and reduce the fitness or capabilities of the organism; or lethal, and kill the cell or organism. Beneficial mutations are rare; much more common are the detrimental mutations that can cause cancer. Where a mutation occurs can also affect how a mutation influences an organism. A mutation to the region of the genome that codes for a given protein will have a major impact on cell or organism behavior. Mutations in other regions of the genome may only have a more indirect impact on the organism.

The most important alternative classification, however, comes in distinguishing whether or not the mutation will potentially be handed down to the organism’s offspring. If the mutation occurs in somatic or general body cells (e.g., skin cells in humans), then the effect of the mutation is limited to the original organism. However, if the mutation occurs in the genetic material that is passed on to offspring, known as the germ line (e.g., egg or sperm cells in humans), effects propagate on to new generations. These propagating changes cause populations to change and evolve. We are the result of the compounding effects of thousands of different mutations to our ancestor’s DNA, determining features from our eye color to our height to our mental capabilities. While we may not possess superpowers, we are all X-Men (or Women) in our own way.

  1. Gabriel, A. & Przybylski, J. (2010) Sickle-cell anemia: A Look at Global Haplotype Distribution. Nature Education 3(3):2
  2. http://dictionary.reference.com/browse/mutation
Posted in Evolution 101 | Tagged , , | Leave a comment