The evolution of learning systems:to Bayes or not to be

Bayesian algorithms pose a limit to the performance learning algorithms can achieve. Natural selection should guide the evolution of information processing systems towards those limits. What can we learn from this evolution and what properties do the intermediate stages have? While this question is too general to permit any answer, progress can be made by restricting the class of information processing systems under study. We present analytical and numerical results for the evolution of on-line algorithms for learning from examples for neural network classifiers, which might include or not a hidden layer. The analytical results are obtained by solving a variational problem to determine the learning algorithm that leads to maximum generalization ability. Simulations using evolutionary programming, for programs that implement learning algorithms, confirm and expand the results. The principal result is not just that the evolution is towards a Bayesian limit. Indeed it is essentially reached. In addition we find that evolution is driven by the discovery of useful structures or combinations of variables and operators. In different runs the temporal order of the discovery of such combinations is unique. The main result is that combinations that signal the surprise brought by an example arise always before combinations that serve to gauge the performance of the learning algorithm. This latter structures can be used to implement annealing schedules. The temporal ordering can be understood analytically as well by doing the functional optimization in restricted functional spaces. We also show that there is data suggesting that the appearance of these traits also follows the same temporal ordering in biological systems. © 2006 American Institute of Physics.

An in-vitro-in-vivo model for the transdermal delivery of cholecalciferol for the purposes of rodent management

The natural selection of anticoagulant resistant rats has resulted in a need for an alternative to anticoagulant rodenticides which differs in both active ingredient and in the method of dosing. Cholecalciferol toxicity to rodents using the dermal route is demonstrated using a variety of penetration enhancing formulations in two in-vitro models and finally in-vivo. A 1 ml dose of 50/50 (v/v) DMSO/ethanol containing 15% (v/v) PEG 200 and 20% (w/v) cholecalciferol was judged as 'sufficiently effective' in line with the European Union's Biocidal Products Regulation (No. 528/2012) during in-vivo studies. This dose was found to cause 100% mortality in a rat population in 64.4 h (±22 h).

A comparative study of life history and gene flow in darters (Pisces: Percidae)

Gene flow, or the exchange of genes between populations, is important because it determines the evolutionary trajectory of a species, including the relative influences of genetic drift and natural selection in the process of population differentiation. Gene flow differs among species because of variation in dispersal capability and abundances across taxa, and historical forces related to geological or lineage history. Both history and ecology influence gene flow in potentially complicated ways, and accounting for their effects remains an important problem in evolutionary biology. This research is a comparative study of gene flow and life-history in a monophyletic group of stream fishes, the darters. As a first step in disentangling historical and ecological effects, I reconstructed the phylogenetic relationships of the study species from nucleotide sequences in the mtDNA control region. I then used this phylogeny and regional glaciation history to infer historical effects on life-history evolution and gene flow in 15 species of darters. Gene flow was estimated indirectly, using information from 20 resolvable and polymorphic allozyme loci. When I accounted for historical effects, comparisons across taxa revealed that gene flow rates were closely associated with differences in clutch sizes and reproductive investment patterns. I hypothesized that differences in larval dispersal among taxa explained this relationship. Results from a field study of larval drift were consistent with this hypothesis. Finally, I asked whether there was an interaction between species' ecology and genetic differentiation across biogeographically distinct regions. Information from allozymes and mtDNA sequences revealed that life history plays an important role in the magnitude of species divergence across biogeographic boundaries. These results suggested an important association between life histories and rates of speciation following an allopatric isolation event. This research, along with other studies from the literature, further illustrates the enormous potential of North American freshwater fishes as a system for studying speciation processes. ^

Social and seasonal plasticity of the electric fish, Brachyhypopomus gauderio

The South American electric knifefish, Brachyhypopomus gauderio, uses weakly electric fields to see and communicate in the dark. Only one study to date has investigated natural behavior in this species during the breeding season; this study proposed that B. guarerio has an exploded lek polygyny breeding system. To test this hypothesis, artificial marshes simulating the native vegetation, temperature, and water conductivities of the South American subtropics were created to study seasonal variation in associative behavior of B. gauderio during the breeding and non-breeding seasons. Mark/recapture methods were used to keep track of individual fish and their dispersion inside the experimental designs. The experimental design proved to be extremely successful at eliciting reproduction. Differences were found in seasonal variations of social behaviors between adult and juvenile populations. Although no apparent sex. differences in movement patterns were found during the breeding season; a trend for male-male aversion was found, suggesting male-male avoidance as a possible strategy guiding aspects of social behaviors in this species. Further, movement may be a tactic for mate seeking as the individuals who moved the most during the breeding season obtained the most opposite sex interactions. These findings support the exploded lek polygyny model. Social interactions are subject to complex regulation by social, physiologic and ecological factors; the extent to which these associations are repeatable may provide novel insights on the evolution of sociality as it has been shaped by natural selection.

Colony-specific calcification and mortality under ocean acidification in the branching coral Montipora digitata

Ocean acidification (OA) threatens calcifying marine organisms including reef-building corals. In this study, we examined the OA responses of individual colonies of the branching scleractinian coral Montipora digitata. We exposed nubbins of unique colonies (n = 15) to ambient or elevated pCO2 under natural light and temperature regimes for 110 days. Although elevated pCO2 exposure on average reduced calcification, individual colonies showed unique responses ranging from declines in positive calcification to negative calcification (decalcification) to no change. Similarly, mortality was greater on average in elevated pCO2, but also showed colony-specific patterns. High variation in colony responses suggests the possibility that ongoing OA may lead to natural selection of OA-tolerant colonies within a coral population.

The Roles of Phenotypic Plasticity and Plant-Microbe Interactions in the Evolution of Complex Traits in Boechera stricta

All organisms live in complex habitats that shape the course of their evolution by altering the phenotype expressed by a given genotype (a phenomenon known as phenotypic plasticity) and simultaneously by determining the evolutionary fitness of that phenotype. In some cases, phenotypic evolution may alter the environment experienced by future generations. This dissertation describes how genetic and environmental variation act synergistically to affect the evolution of glucosinolate defensive chemistry and flowering time in Boechera stricta, a wild perennial herb. I focus particularly on plant-associated microbes as a part of the plant’s environment that may alter trait evolution and in turn be affected by the evolution of those traits. In the first chapter I measure glucosinolate production and reproductive fitness of over 1,500 plants grown in common gardens in four diverse natural habitats, to describe how patterns of plasticity and natural selection intersect and may influence glucosinolate evolution. I detected extensive genetic variation for glucosinolate plasticity and determined that plasticity may aid colonization of new habitats by moving phenotypes in the same direction as natural selection. In the second chapter I conduct a greenhouse experiment to test whether naturally-occurring soil microbial communities contributed to the differences in phenotype and selection that I observed in the field experiment. I found that soil microbes cause plasticity of flowering time but not glucosinolate production, and that they may contribute to natural selection on both traits; thus, non-pathogenic plant-associated microbes are an environmental feature that could shape plant evolution. In the third chapter, I combine a multi-year, multi-habitat field experiment with high-throughput amplicon sequencing to determine whether B. stricta-associated microbial communities are shaped by plant genetic variation. I found that plant genotype predicts the diversity and composition of leaf-dwelling bacterial communities, but not root-associated bacterial communities. Furthermore, patterns of host genetic control over associated bacteria were largely site-dependent, indicating an important role for genotype-by-environment interactions in microbiome assembly. Together, my results suggest that soil microbes influence the evolution of plant functional traits and, because they are sensitive to plant genetic variation, this trait evolution may alter the microbial neighborhood of future B. stricta generations. Complex patterns of plasticity, selection, and symbiosis in natural habitats may impact the evolution of glucosinolate profiles in Boechera stricta.

Seawater carbonate chemistry and Celleporella hyalina biological processes during experiments, 2011

Increased anthropogenic CO2 emissions in the last two centuries have lead to rising sea surface temperature and falling ocean pH, and it is predicted that current global trends will worsen over the next few decades. There is limited understanding of how genetic variation among individuals will influence the responses of populations and species to these changes. A microcosm system was set up to study the effects of predicted temperature and CO2 levels on the bryozoan Celleporella hyalina. In this marine species, colonies grow by the addition of male, female and feeding modular individuals (zooids) and can be physically subdivided to produce a clone of genetically identical colonies. We studied colony growth rate (the addition of zooids), reproductive investment (the ratio of sexual to feeding zooids) and sex ratio (male to female zooids) in four genetically distinct clonal lines. There was a significant effect of clone on growth rate, reproductive investment and sex ratio, with clones showing contrasting responses to the various temperature and pH combinations. Overall, decreasing pH and increasing temperature caused reduction of growth, and eventual cessation of growth was often observed at the highest temperature, especially during the latter half of the 15-day trials. Reproductive investment increased with increasing temperature and decreasing pH, varying more widely with temperature at the lowest pH. The increased production of males, a general stress response of the bryozoan, was seen upon exposure to reduced pH, but was not expressed at the highest temperature tested, presumably due to the frequent cessation of growth. Further to the significant effect of pH on the measured whole-colony parameters, observation by scanning electron microscopy revealed surface pitting of the calcified exoskeleton in colonies that were exposed to increased acidity. Studying ecologically relevant processes of growth and reproduction, we demonstrate the existence of relevant levels of variation among genetic individuals which may enable future adaptation via non-mutational natural selection to falling pH and rising temperature.

The onset of lactase persistence in Europe

The genomic region containing the lactase (LCT) gene shows one of the strongest signals of positive selection in Europeans, detectable using a range of approaches including haplotype length, linked microsatellite variation and population-differentiation-based tests. Lactase is the enzyme that carries out the digestion of the milk sugar lactose. Its expression decreases at some point after the weaning period is over in most mammals and in around 68% of all living adult humans. However, in some humans, particularly those from populations with a history of dairying, lactase is expressed throughout adulthood. This trait is called lactase persistence (LP), and in people of European ancestry, it is associated with a single mutation (-13910*T). Evidence from the detection of dairy fat residues in potsherds, and allele frequencies in ancient DNA samples suggest that LP arose after dairying practices had developed. However, the reasons why LP may have been advantageous are still debated, and the respective contribution of demography and natural selection remains to be disentangled. This paper discusses various studies, from archaeology to population genetics, that have shed some light on the subject by investigating the evolution of LP in Europe.

The Quantum Mitochondrion and Optimal Health

A sufficiently complex set of molecules, if subject to perturbation, will self-organise and show emergent behaviour. If such a system can take on information it will become subject to natural selection. This could explain how self-replicating molecules evolved into life and how intelligence arose. A pivotal step in this evolutionary process was of course the emergence of the eukaryote and the advent of the mitochondrion, which both enhanced energy production per cell and increased the ability to process, store and utilise information. Recent research suggest that from its inception life embraced quantum effects such as “tunnelling” and “coherence” while competition and stressful conditions provided a constant driver for natural selection. We believe that the biphasic adaptive response to stress described by hormesis – a process that captures information to enable adaptability, is central to this whole process. Critically, hormesis could improve mitochondrial quantum efficiency, improving the ATP/ROS ratio, while inflammation, which is tightly associated with the aging process, might do the opposite. This all suggests that to achieve optimal health and healthy ageing, one has to sufficiently stress the system to ensure peak mitochondrial function, which itself could reflect selection of optimum efficiency at the quantum level.

A Biocultural Examination of Student Learning Behaviors in Large Undergraduate Lectures

Thesis (Ph.D.)--University of Washington, 2016-08

Phocid seal leptin: tertiary structure and hydrophobic receptor binding site preservation during distinct leptin gene evolution

The cytokine hormone leptin is a key signalling molecule in many pathways that control physiological functions. Although leptin demonstrates structural conservation in mammals, there is evidence of positive selection in primates, lagomorphs and chiropterans. We previously reported that the leptin genes of the grey and harbour seals (phocids) have significantly diverged from other mammals. Therefore we further investigated the diversification of leptin in phocids, other marine mammals and terrestrial taxa by sequencing the leptin genes of representative species. Phylogenetic reconstruction revealed that leptin diversification was pronounced within the phocid seals with a high dN/dS ratio of 2.8, indicating positive selection. We found significant evidence of positive selection along the branch leading to the phocids, within the phocid clade, but not over the dataset as a whole. Structural predictions indicate that the individual residues under selection are away from the leptin receptor (LEPR) binding site. Predictions of the surface electrostatic potential indicate that phocid seal leptin is notably different to other mammalian leptins, including the otariids. Cloning the grey seal leptin binding domain of LEPR confirmed that this was structurally conserved. These data, viewed in toto, support a hypothesis that phocid leptin divergence is unlikely to have arisen by random mutation. Based upon these phylogenetic and structural assessments, and considering the comparative physiology and varying life histories among species, we postulate that the unique phocid diving behaviour has produced this selection pressure. The Phocidae includes some of the deepest diving species, yet have the least modified lung structure to cope with pressure and volume changes experienced at depth. Therefore, greater surfactant production is required to facilitate rapid lung re-inflation upon surfacing, while maintaining patent airways. We suggest that this additional surfactant requirement is met by the leptin pulmonary surfactant production pathway which normally appears only to function in the mammalian foetus.

Competition and fixation of cohorts of adaptive mutations under Fisher geometrical model

One of the simplest models of adaptation to a new environment is Fisher's Geometric Model (FGM), in which populations move on a multidimensional landscape defined by the traits under selection. The predictions of this model have been found to be consistent with current observations of patterns of fitness increase in experimentally evolved populations. Recent studies investigated the dynamics of allele frequency change along adaptation of microbes to simple laboratory conditions and unveiled a dramatic pattern of competition between cohorts of mutations, i.e., multiple mutations simultaneously segregating and ultimately reaching fixation. Here, using simulations, we study the dynamics of phenotypic and genetic change as asexual populations under clonal interference climb a Fisherian landscape, and ask about the conditions under which FGM can display the simultaneous increase and fixation of multiple mutations-mutation cohorts-along the adaptive walk. We find that FGM under clonal interference, and with varying levels of pleiotropy, can reproduce the experimentally observed competition between different cohorts of mutations, some of which have a high probability of fixation along the adaptive walk. Overall, our results show that the surprising dynamics of mutation cohorts recently observed during experimental adaptation of microbial populations can be expected under one of the oldest and simplest theoretical models of adaptation-FGM.

Drosophila Adaptation to Viral Infection through Defensive Symbiont Evolution

Microbial symbionts can modulate host interactions with biotic and abiotic factors. Such interactions may affect the evolutionary trajectories of both host and symbiont. Wolbachia protects Drosophila melanogaster against several viral infections and the strength of the protection varies between variants of this endosymbiont. Since Wolbachia is maternally transmitted, its fitness depends on the fitness of its host. Therefore, Wolbachia populations may be under selection when Drosophila is subjected to viral infection. Here we show that in D. melanogaster populations selected for increased survival upon infection with Drosophila C virus there is a strong selection coefficient for specific Wolbachia variants, leading to their fixation. Flies carrying these selected Wolbachia variants have higher survival and fertility upon viral infection when compared to flies with the other variants. These findings demonstrate how the interaction of a host with pathogens shapes the genetic composition of symbiont populations. Furthermore, host adaptation can result from the evolution of its symbionts, with host and symbiont functioning as a single evolutionary unit.

Phylogenetic inference's algorithms

Phylogenetic inference consist in the search of an evolutionary tree to explain the best way possible genealogical relationships of a set of species. Phylogenetic analysis has a large number of applications in areas such as biology, ecology, paleontology, etc. There are several criterias which has been defined in order to infer phylogenies, among which are the maximum parsimony and maximum likelihood. The first one tries to find the phylogenetic tree that minimizes the number of evolutionary steps needed to describe the evolutionary history among species, while the second tries to find the tree that has the highest probability of produce the observed data according to an evolutionary model. The search of a phylogenetic tree can be formulated as a multi-objective optimization problem, which aims to find trees which satisfy simultaneously (and as much as possible) both criteria of parsimony and likelihood. Due to the fact that these criteria are different there won't be a single optimal solution (a single tree), but a set of compromise solutions. The solutions of this set are called "Pareto Optimal". To find this solutions, evolutionary algorithms are being used with success nowadays.This algorithms are a family of techniques, which aren’t exact, inspired by the process of natural selection. They usually find great quality solutions in order to resolve convoluted optimization problems. The way this algorithms works is based on the handling of a set of trial solutions (trees in the phylogeny case) using operators, some of them exchanges information between solutions, simulating DNA crossing, and others apply aleatory modifications, simulating a mutation. The result of this algorithms is an approximation to the set of the “Pareto Optimal” which can be shown in a graph with in order that the expert in the problem (the biologist when we talk about inference) can choose the solution of the commitment which produces the higher interest. In the case of optimization multi-objective applied to phylogenetic inference, there is open source software tool, called MO-Phylogenetics, which is designed for the purpose of resolving inference problems with classic evolutionary algorithms and last generation algorithms. REFERENCES [1] C.A. Coello Coello, G.B. Lamont, D.A. van Veldhuizen. Evolutionary algorithms for solving multi-objective problems. Spring. Agosto 2007 [2] C. Zambrano-Vega, A.J. Nebro, J.F Aldana-Montes. MO-Phylogenetics: a phylogenetic inference software tool with multi-objective evolutionary metaheuristics. Methods in Ecology and Evolution. En prensa. Febrero 2016.