Numerical analysis of a reinforcement learning model with the dynamic aspiration level in the iterated Prisoner's dilemma

The phenomenon of geographic parthenogenesis is closely tied to the question of why sexual reproduction is the dominant mode of reproduction in animals and plants. Geographic parthenogenesis describes the fact that many species reproduce asexually at the boundaries of their range. We present a mathematical model that derives the dominance of sexuals at the center and the dominance of asexuals at the boundary of a species' range from exactly the same mechanism. Our model is based on a set of resources that regrow slowly and that can be consumed only by those individuals that have a suitable genotype. Genotype is implemented by a multilocus model with two alleles at each locus, and with free recombination during production of sexual offspring. The model is tailored to seasonal species with intermittent mixis and low survival of offspring, such as Daphnia and aphids. Several patches of resources are arranged in a row, with a gradient of those parameters that typically vary through the range of species. By letting sexually and asexually reproducing populations compete, we obtain the typical patterns of geographic parthenogenesis.     

Ubiquitous crossmodal Stochastic Resonance in humans: auditory noise facilitates tactile, visual and proprioceptive sensations

Background

Stochastic resonance is a nonlinear phenomenon whereby the addition of noise can improve the detection of weak stimuli. An optimal amount of added noise results in the maximum enhancement, whereas further increases in noise intensity only degrade detection or information content. The phenomenon does not occur in linear systems, where the addition of noise to either the system or the stimulus only degrades the signal quality. Stochastic Resonance (SR) has been extensively studied in different physical systems. It has been extended to human sensory systems where it can be classified as unimodal, central, behavioral and recently crossmodal. However what has not been explored is the extension of this crossmodal SR in humans. For instance, if under the same auditory noise conditions the crossmodal SR persists among different sensory systems.

Methodology/Principal Findings

Using physiological and psychophysical techniques we demonstrate that the same auditory noise can enhance the sensitivity of tactile, visual and propioceptive system responses to weak signals. Specifically, we show that the effective auditory noise significantly increased tactile sensations of the finger, decreased luminance and contrast visual thresholds and significantly changed EMG recordings of the leg muscles during posture maintenance.

Conclusions/Significance

We conclude that crossmodal SR is a ubiquitous phenomenon in humans that can be interpreted within an energy and frequency model of multisensory neurons spontaneous activity. Initially the energy and frequency content of the multisensory neurons'' activity (supplied by the weak signals) is not enough to be detected but when the auditory noise enters the brain, it generates a general activation among multisensory neurons of different regions, modifying their original activity. The result is an integrated activation that promotes sensitivity transitions and the signals are then perceived. A physiologically plausible model for crossmodal stochastic resonance is presented.     

Influence of simulated neuromuscular noise on the dynamic stability and fall risk of a 3D dynamic walking model

Measures that can predict risk of falling are essential for enrollment of older adults into fall prevention programs. Local and orbital stability directly quantify responses to very small perturbations and are therefore putative candidates for predicting fall risk. However, research to date is not conclusive on whether and how these measures relate to fall risk. Testing this empirically would be time consuming or may require high risk tripping experiments. Simulation studies therefore provide an important tool to initially explore potential measures to predict fall risk. This study performed simulations with a 3D dynamic walking model to explore if and how dynamic stability measures predict fall risk. The model incorporated a lateral step controller to maintain lateral stability. Neuronal noise of increasing amplitude was added to this controller to manipulate fall risk. Short-term (λ(S)(*)) local instability did predict fall risk, but long-term (λ(L)(*)) local instability and orbital stability (maxFM) did not. Additionally, λ(S)(*) was an early predictor for fall risk as it started increasing before fall risk increased. Therefore, λ(S)(*) could be a very useful tool to identify older adults whose fall risk is about to increase, so they can be enrolled in fall prevention programs before they actually fall.     

Stochastic noise interferes coherently with a model biological clock and produces specific dynamic behaviour.

The influence of noise is unavoidable in all living systems. Its impact on a model of a biological clock, normally running in regular oscillating modes, is examined. It is shown that in a specific system in which endogenous rhythmicity is produced by a beat oscillator acting on a feedback coupled metabolic pool system, noise can act coherently to produce unexpected dynamic behaviour, running from regular over pseudo-regular to irregular time-structures. If the biological system consists of a set of identical weakly coupled cells, stochasticity may lead to phase decoupling producing irregular spatio-temporal patterns. Synchronization via phase resetting can be achieved by external short-time temperature pulses. Explicit results are obtained for the well-studied circadian photosynthesis oscillations in plants performing crassulacean acid metabolism. Because of the generic structure of the underlying nonlinear dynamics they can, however, be regarded as a general property of the influence of noise on nonlinear excitable systems with fixed points occuring close to limit cycles.     

Encoding of relative enclosure size in a dynamic three-dimensional virtual environment by humans

Human participants searched in a dynamic three-dimensional virtual-environment rectangular enclosure for a distinctly colored bin located in one of the four corners. During test trials, all bins were rendered identical in color, and the shape of the rectangular search space either remained the same or was modified to a relatively sized contracted rectangle, an expanded rectangle, or a square. Participants made one choice response during test trials. In the rectangular enclosures, more of participants’ choice responses were allocated to the geometrically correct corners than to the geometrically incorrect corners. In the square enclosure, participants’ choice responses were allocated equivalently to each of the four corners. Results replicate previous enclosure size studies demonstrating encoding of enclosure geometry with human and non-human animal subjects conducted in real environments and extend these results to include encoding of relative enclosure geometry. Results are discussed with respect to theoretical accounts of geometry learning.     

Influence of simulated neuromuscular noise on movement variability and fall risk in a 3D dynamic walking model

People at risk of falling exhibit increased gait variability, which may predict future falls. However, the causal mechanisms underlying these correlations are not well known. Increased neuronal noise associated with aging likely leads to increased gait variability, which could in turn lead to increased fall risk. This paper presents a model of how changes in neuromuscular noise independently affect gait variability and probability of falling, and aims to determine the extent to which changes in gait variability directly predict fall risk. We used a dynamic walking model that incorporates a lateral step controller to maintain lateral stability. Noise was applied to this controller to approximate neuromuscular noise in humans. Noise amplitude was varied between low amplitudes that did not induce falls and high amplitudes for which the model always fell. With increases in noise amplitude, the model fell more often and after fewer steps. Gait variability increased with noise amplitude and predicted increased probability of falling. Importantly, these relationships were not linear. At either low gait variability or very high gait variability, small increases in noise and variability affected probability of falling very little. Conversely, at intermediate noise and/or variability levels, the same small increases resulted in large increases in probability of falling. Our results validate the idea that age-related increases in neuromuscular noise likely play a direct contributing role in increasing fall risk. However, neuromuscular noise remains only one of many important factors that need to be considered. These findings have important implications for fall prevention research and practice.     

Theoretical analysis of a TIME-FREQUENCY-PCNN auditory cortex model

A particular pulsed neural model of the auditory cortex, the Time-Frequency Pulse Coupled Neural Network (TF-PCNN), has shown to decompose its stimulus from the cochlea into characteristic pulse-coded time-frequency (TF) stop and pass regions. A derived sum of spectrograms representation with these zero- or one-valued TF weightings has already been applied for the denoising of speech in a previous work. This decomposition is now related to the concept of TF projection filters, which allows to reinterpret the model equations. The functionality imposed by the model equations can so be accessed and justified from a TF signal processing perspective, in addition to its biological motivation from experimentally observed neurophysiological behavior and simulations.     

Evolution of a rippled membrane during phospholipase A2 hydrolysis studied by time-resolved AFM

The sensitivity of phospholipase A(2) (PLA(2)) for lipid membrane curvature is explored by monitoring, through time-resolved atomic force microscopy, the hydrolysis of supported double bilayers in the ripple phase. The ripple phase presents a corrugated morphology. PLA(2) is shown to have higher activity toward the ripple phase compared to the gel phase in 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) membranes, indicating its preference for this highly curved membrane morphology. Hydrolysis of the stable and metastable ripple structures is monitored for equimolar DMPC/1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC)-supported double bilayers. As shown by high-performance liquid chromatography results, DSPC is resistant to hydrolysis at this temperature, resulting in a more gradual hydrolysis of the surface that leads to a change in membrane morphology without loss of membrane integrity. This is reflected in an increase in ripple spacing, followed by a sudden flattening of the lipid membrane during hydrolysis. Hydrolysis of the ripple phase results in anisotropic holes running parallel to the ripples, suggesting that the ripple phase has strip regions of higher sensitivity to enzymatic attack. Bulk high-performance liquid chromatography measurements indicate that PLA(2) preferentially hydrolyzes DMPC in the DMPC/DSPC ripples. We suggest that this leads to the formation of a flat gel-phase lipid membrane due to enrichment in DSPC. The results point to the ability of PLA(2) for inducing a compositional phase transition in multicomponent membranes through preferential hydrolysis while preserving membrane integrity.     

Biosensor detection of triplex formation by modified oligonucleotides

Due to the instability of DNA oligonucleotides in biological solutions, antisense or antigene therapies aimed at modulation of specific gene expression will most likely require the use of oligonucleotides with modified backbones. Here, we examine the use of a surface plasmon resonance biosensor (BIAcore) to compare triplex-directed binding of modified oligonucleotides targeted to a region of the murine c-myc promoter. We describe optimization of experimental conditions to minimize nonspecific interactions between the oligonucleotides and the sensor chip surface, and the limitations imposed by certain backbones and sequence types. The abilities of pyrimidine oligonucleotides with various modified backbones to form specific triple helices with an immobilized hairpin duplex were readily determined using the biosensor. Modification of the third-strand oligonucleotide with RNA or 2(')-O-methyl RNA was found to enhance triplex formation, whereas phosphorothioate or phosphotriester substitutions abrogated it. A comparison of these results to DNase I footprinting experiments using the same oligonucleotides showed complete agreement between the two sets of data.     

A DNA solution of SAT problem by a modified sticker model

Among various DNA computing algorithms, it is very common to create an initial data pool that covers correct and incorrect answers at first place followed by a series of selection process to destroy the incorrect ones. The surviving DNA sequences are read as the solutions to the problem. However, algorithms based on such a brute force search will be limited to the problem size. That is, as the number of parameters in the studied problem grows, eventually the algorithm becomes impossible owing to the tremendous initial data pool size. In this theoretical work, we modify a well-known sticker model to design an algorithm that does not require an initial data pool for SAT problem. We propose to build solution sequences in parts to satisfy one clause in a step, and eventually solve the whole Boolean formula after a number of steps. Accordingly, the size of data pool grows from one sort of molecule to the number of solution assignments. The proposed algorithm is expected to provide a solution to SAT problem and become practical as the problem size scales up.     

Synergistic mixture interactions in detection of perithreshold odors by humans

Laboratory demonstrations of synergistic mixture interactions in human odor perception have been rare. The current study examined perithreshold mixture interactions between maple lactone (ML) and selected carboxylic acids. An air-dilution olfactometer allowed precise stimulus control. Experimenters measured stimulus concentrations in vapor phase using a combination of solid-phase microextraction and gas chromatography/mass spectrometry. A probability of detection versus concentration, or a psychometric, functions was measured for pure ML. Psychometric functions were also measured for ML with the addition of fixed, subthreshold concentrations of carboxylic acids. Relative to statistical independence in detection, clear synergy occurred over a range of ML concentrations. To the best of our knowledge, the current results constitute the first clear demonstration of synergy in odor detection by humans from an experiment that combined precise stimulus control, vapor-phase calibration of stimuli, and a clear statistical definition of synergy.     

Demographic noise and resilience in a semi-arid ecosystem model

The scarcity of water characterising drylands forces vegetation to adopt appropriate survival strategies. Some of these generate water–vegetation feedback mechanisms that can lead to spatial self-organisation of vegetation, as it has been shown with models representing plants by a density of biomass, varying continuously in time and space. However, although plants are usually quite plastic they also display discrete qualities and stochastic behaviour. These features may give rise to demographic noise, which in certain cases can influence the qualitative dynamics of ecosystem models. In the present work we explore the effects of demographic noise on the resilience of a model semi-arid ecosystem. We introduce a spatial stochastic eco-hydrological hybrid model in which plants are modelled as discrete entities subject to stochastic dynamical rules, while the dynamics of surface and soil water are described by continuous variables. The model has a deterministic approximation very similar to previous continuous models of arid and semi-arid ecosystems. By means of numerical simulations we show that demographic noise can have important effects on the extinction and recovery dynamics of the system. In particular we find that the stochastic model escapes extinction under a wide range of conditions for which the corresponding deterministic approximation predicts absorption into desert states.     

Pitch change of a continuous tone activates two distinct processes in human auditory cortex: a study with whole-head magnetometer

Previous studies have shown that a frequency change in a continuous tone elicits an NI type of ERP (event-related potential) component. It remained unclear, however, whether this response is a “genuine” N1 (onset detector response) or the mismatch negativity (MMN), a change-detector type of ERP response, elicited in previous studies by an infrequent change in a sequence of homogeneous stimuli. A further possibility is a nearly perfect overlap of the two types of ERP components. The advent of modern, high-resolution magnetometers has opened a new, powerful way to tackle such component-overlap problems.Subjects were presented with a continuous tone of 988 Hz which was occasionally increased to 1108 Hz for a period of 100 msec. The magnetic responses to this change consisted of two partially overlapping components with peaks separated by 30 msec. The earlier component was probably generated by neuronal populations of the auditory cortex corresponding to the supratemporal N1, whereas the later one, generated anteriorly and inferiorly to the first, probably reflects a mismatch process causing the magnetic equivalent of the electrical MMN.     

The evolution of punishment and apology: an iterated prisoner's dilemma model

Recently, behaviors that seem to function as punishment or apology have been reported among non-human primates as well as humans. Such behaviors appear to play an important role in maintaining cooperation between individuals. Therefore, the evolution of these behaviors should be examined from the viewpoint of the evolution of cooperation. The iterated prisoner's dilemma (IPD) game is generally considered to be a standard model for the evolution of cooperation. In the present study, strategies accompanied by punishment-like attacks or apology-like behavior were introduced into the common IPD simulation. Punishment and apology were represented by the P signal and the AS signal given immediately after defection. A strategy with the P and AS signals, named the pPAS strategy, was proved to be an evolutionarily stable strategy under certain conditions. Numerical simulations were carried out according to different assigned values of the costs of punishment and apology. The simulations showed that pPAS could dominate the population (1) when the cost of giving P is relatively small, (2) when the cost of receiving P is relatively large, or (3) when the cost of giving AS is relatively large. The relative cost of giving AS had the clearest effect on the success of pPAS. pPAS can dominate the population even when a dominance asymmetry of the costs between two players was introduced. The present results suggest the possible evolution of social behaviors like punishment or apology as a means of maintaining cooperation. This revised version was published online in November 2006 with corrections to the Cover Date.     

Convergence properties of a modified Hopfield-Tank model

The neural network model of Hopfield and Tank applied to the Travelling Salesman Problem, has been analyzed in order to improve its convergence properties. A simple change of the parameter sets always allows to reach states corresponding to valid tours. Besides a more interesting modification has been presented by adding a new term to force expression. This modified model has a high value of convergence and it is able to find short tours. So, more confidence can be given to these type of models, and real applications could be performed.     

Requirements for cleavage by a modified RNase P of a small model substrate.

M1 RNA, the catalytic RNA subunit of RNase P from Escherichia coli, has been covalently linked at its 3' terminus to oligonucleotides (guide sequences) that guide the enzyme to target RNAs through hybridization with the target sequences. These constructs (M1GS RNAs) have been used to determine some minimal features of model substrates. As few as 3 bp on the 3' side of the site of cleavage in a substrate complex and 1 nt on the 5' side are required for cleavage to occur. The cytosines in the 3' terminal CCA sequence of the model substrates are important for cleavage efficiency but not cleavage site selection. A purine (base-paired or not) at the 3' side of the cleavage site is important both for cleavage site selection and efficiency. M1GS RNAs provide both a simple system for characterization of the reaction governed by M1 RNA and a tool for gene therapy.