Weak Selection Research Articles

HybridSNN: Combining Bio-Machine Strengths by Boosting Adaptive Spiking Neural Networks.

Spiking neural networks (SNNs), inspired by the neuronal network in the brain, provide biologically relevant and low-power consuming models for information processing. Existing studies either mimic the learning mechanism of brain neural networks as closely as possible, for example, the temporally local learning rule of spike-timing-dependent plasticity (STDP), or apply the gradient descent rule to optimize a multilayer SNN with fixed structure. However, the learning rule used in the former is local and how the real brain might do the global-scale credit assignment is still not clear, which means that those shallow SNNs are robust but deep SNNs are difficult to be trained globally and could not work so well. For the latter, the nondifferentiable problem caused by the discrete spike trains leads to inaccuracy in gradient computing and difficulties in effective deep SNNs. Hence, a hybrid solution is interesting to combine shallow SNNs with an appropriate machine learning (ML) technique not requiring the gradient computing, which is able to provide both energy-saving and high-performance advantages. In this article, we propose a HybridSNN, a deep and strong SNN composed of multiple simple SNNs, in which data-driven greedy optimization is used to build powerful classifiers, avoiding the derivative problem in gradient descent. During the training process, the output features (spikes) of selected weak classifiers are fed back to the pool for the subsequent weak SNN training and selection. This guarantees HybridSNN not only represents the linear combination of simple SNNs, as what regular AdaBoost algorithm generates, but also contains neuron connection information, thus closely resembling the neural networks of a brain. HybridSNN has the benefits of both low power consumption in weak units and overall data-driven optimizing strength. The network structure in HybridSNN is learned from training samples, which is more flexible and effective compared with existing fixed multilayer SNNs. Moreover, the topological tree of HybridSNN resembles the neural system in the brain, where pyramidal neurons receive thousands of synaptic input signals through their dendrites. Experimental results show that the proposed HybridSNN is highly competitive among the state-of-the-art SNNs.

Environmental variance in male mating success modulates the positive versus negative impacts of sexual selection on genetic load.

Sexual selection on males is predicted to increase population fitness, and delay population extinction, when mating success negatively covaries with genetic load across individuals. However, such benefits of sexual selection could be counteracted by simultaneous increases in genome-wide drift resulting from reduced effective population size caused by increased variance in fitness. Resulting fixation of deleterious mutations could be greatest in small populations, and when environmental variation in mating traits partially decouples sexual selection from underlying genetic variation. The net consequences of sexual selection for genetic load and population persistence are therefore likely to be context dependent, but such variation has not been examined. We use a genetically explicit individual-based model to show that weak sexual selection can increase population persistence time compared to random mating. However, for stronger sexual selection such positive effects can be overturned by the detrimental effects of increased genome-wide drift. Furthermore, the relative strengths of mutation-purging and drift critically depend on the environmental variance in the male mating trait. Specifically, increasing environmental variance caused stronger sexual selection to elevate deleterious mutation fixation rate and mean selection coefficient, driving rapid accumulation of drift load and decreasing population persistence times. These results highlight an intricate balance between conflicting positive and negative consequences of sexual selection on genetic load, even in the absence of sexually antagonistic selection. They imply that environmental variances in key mating traits, and intrinsic genetic drift, should be properly factored into future theoretical and empirical studies of the evolution of population fitness under sexual selection.

The Complex Proteolipidic Behavior of the SARS-CoV-2 Envelope Protein Channel: Weak Selectivity and Heterogeneous Oligomerization.

The envelope (E) protein is a small polypeptide that can form ion channels in coronaviruses. In SARS coronavirus 2 (SARS-CoV-2), the agent that caused the recent COVID-19 pandemic, and its predecessor SARS-CoV-1, E protein is found in the endoplasmic reticulum-Golgi intermediate compartment (ERGIC), where virion budding takes place. Several reports claim that E protein promotes the formation of "cation-selective channels". However, whether this term represents specificity to certain ions (e.g., potassium or calcium) or the partial or total exclusion of anions is debatable. Herein, we discuss this claim based on the available data for SARS-CoV-1 and -2 E and on new experiments performed using the untagged full-length E protein from SARS-CoV-2 in planar lipid membranes of different types, including those that closely mimic the ERGIC membrane composition. We provide evidence that the selectivity of the E-induced channels is very mild and depends strongly on lipid environment. Thus, despite past and recent claims, we found no indication that the E protein forms cation-selective channels that prevent anion transport, and even less that E protein forms bona fide specific calcium channels. In fact, the E channel maintains its multi-ionic non-specific neutral character even in concentrated solutions of Ca2+ ions. Also, in contrast to previous studies, we found no evidence that SARS-CoV-2 E channel activation requires a particular voltage, high calcium concentrations or low pH, in agreement with available data from SARS-CoV-1 E. In addition, sedimentation velocity experiments suggest that the E channel population is mostly pentameric, but very dynamic and probably heterogeneous, consistent with the broad distribution of conductance values typically found in electrophysiological experiments. The latter has been explained by the presence of proteolipidic channel structures.

Alkali metal cations modulate the geometry of different binding sites in HCN4 selectivity filter for permeation or block.

Hyperpolarization-activated cyclic-nucleotide gated (HCN) channels are important for timing biological processes like heartbeat and neuronal firing. Their weak cation selectivity is determined by a filter domain with only two binding sites for K+ and one for Na+. The latter acts as a weak blocker, which is released in combination with a dynamic widening of the filter by K+ ions, giving rise to a mixed K+/Na+ current. Here, we apply molecular dynamics simulations to systematically investigate the interactions of five alkali metal cations with the filter of the open HCN4 pore. Simulations recapitulate experimental data like a low Li+ permeability, considerable Rb+ conductance, a block by Cs+ as well as a punch through of Cs+ ions at high negative voltages. Differential binding of the cation species in specific filter sites is associated with structural adaptations of filter residues. This gives rise to ion coordination by a cation-characteristic number of oxygen atoms from the filter backbone and solvent. This ion/protein interplay prevents Li+, but not Na+, from entry into and further passage through the filter. The site equivalent to S3 in K+ channels emerges as a preferential binding and presumably blocking site for Cs+. Collectively, the data suggest that the weak cation selectivity of HCN channels and their block by Cs+ are determined by restrained cation-generated rearrangements of flexible filter residues.

A green flotation collector towards efficient separation of chalcopyrite from pyrite with high selectivity

Xanthate, due to its superior collecting ability, is widely used in the field of minerals flotation. However, its weak selectivity and foul odor make it difficult to separate chalcopyrite from pyrite and reduce pollution. This work provides an efficient and odorless collector to selectively separate chalcopyrite from pyrite. Amide group was introduced into xanthate and an efficient and odorless xanthate derivative was designed, potassium O-(6-(butylamino)-6-oxohexyl) xanthate (PBAHX), to selectively separate chalcopyrite and pyrite. The flotation results showed that PBAHX exhibited superior selectivity, the separation index of PBAHX for chalcopyrite and pyrite is 2.65 while that of the traditional sodium isobutyl xanthate (SIBX) is 1.29, and enabled the separation of chalcopyrite and pyrite. The electronic-nose detection demonstrated that the odor intensity of PBAHX was lower than SIBX. Subsequently, “horizontal” adsorption model was formed for the adsorption of PBAHX on pyrite to explain the differences in flotation results. Besides, hydrogen bonds may be produced between collector molecules to dampen odor. PBAHX shows potential as a green and effective alternative to sulfide mineral flotation.

SCWOMP Recovery Algorithm for 5G MIMO Communication Symbol Detection

In order to solve the problem of small capacity and high energy consumption in China’s 5G communication technology system, the research proposes that based on the segmented weakly orthogonal matching pursuit (SWOMP) algorithm, it is combined with the compressed sensing matching pursuit algorithm to form a segmented backtracking weak selection positive algorithm and Cross Match Tracking (SCWOMP) algorithm. First, the sparseness of MIMO system technology and its transmission structure is analyzed. Then, the new model is built after comparing with other algorithms, and the problem of overestimating the low recovery probability in the calculation process is improved by the backtracking of the algorithm and the improvement of the angle of the atomic column selection, so as to reduce the number of iterations and improve the performance of the algorithm. The results show that, in the performance comparison of different sampling points under different compressed sensing recovery algorithms, the recovery probability of the SCWOMP algorithm is the best, and when the number of sampling points is 80, although the fixed step size of the SCWOMP algorithm is different, there is recovery. The probability has a maximum value, close to 1. Then, the improved compressed sensing recovery algorithm is simulated and analyzed. When the pruning coefficient is 0.5 and the number of sampling points is 80, the reconstruction rate has a maximum value, and when other algorithms reach the maximum reconstruction rate, the number of sampling points (M) is significantly greater than that of the SCWOMP algorithm. An increase in the rate of reduction of the reconstruction probability of the SCWOMP algorithm is significantly lower than that of other algorithms; when sparsity is equal to 70, the reconstruction probability becomes 0, indicating that SCWOMP has a wider reconfigurable range and has a significant performance effect. This shows that the proposed SCWOMP algorithm has the best detection performance for 5G communication symbol detection, which can effectively increase the capacity of the system and better promote technology.

Complete Mitochondrial Genome Sequence and Phylogenetic Analysis of Procambarus clarkii and Cambaroides dauricus from China.

To enhance the management and protection of crayfish genetic diversity and germplasm resources in Cambaroides dauricus (C. dauricus), a common species of Procambarus clarkii (P. clarkii) was used as a control group to compare the whole mitochondrial genome sequence using Illumina sequencing technology. This study found that the mitochondrial genome of C. dauricus is 15580 bp in length, with a base composition of A (31.84%), G (17.66%), C (9.42%), and T (41.08%) and a C + G content of 27.08%. The C + G in the D-loop is rich in 17.06%, indicating a significant preference. The mitochondrial genome of C. dauricus contains 13 protein-coding genes, 22 tRNA genes, and 2 rRNA genes, with most of the genes labeled in the negative direction, except for a few genes that are labeled in the positive direction. The start codons of the ten coding sequences are ATG, and the quintessential TAA and TAG are the stop codons. This study also found that the Ka/Ks ratios of most protein-coding genes in the mitochondria of both shrimps are lower than 1, indicating weak natural selection, except for nad 2, nad 5, and cox 1. The Ka/Ks ratio of cox 3 is the lowest (less than 0.1), indicating that this protein-coding gene bears strong natural selection pressure and functional constraint in the process of mitochondrial genetic evolution of both shrimps. Furthermore, we constructed phylogenetic analyses based on the entire sequence, which effectively distinguishes the high body from other shrimp species of the genus based on the mitochondrial genome. This study provides molecular genetic data for the diversity investigation and protection of fishery resources with Chinese characteristics and a scientific reference for the evolutionary study of Procambarus.

Characterizing the evolution and phenotypic impact of ampliconic Y chromosome regions

A major part of the human Y chromosome consists of palindromes with multiple copies of genes primarily expressed in testis, many of which have been claimed to affect male fertility. Here we examine copy number variation in these palindromes based on whole genome sequence data from 11,527 Icelandic men. Using a subset of 7947 men grouped into 1449 patrilineal genealogies, we infer 57 large scale de novo copy number mutations affecting palindrome 1. This corresponds to a mutation rate of 2.34 × 10−3 mutations per meiosis, which is 4.1 times larger than our phylogenetic estimate of the mutation rate (5.72 × 10−4), suggesting that de novo mutations on the Y are lost faster than expected under neutral evolution. Although simulations indicate a selection coefficient of 1.8% against non-reference copy number carriers, we do not observe differences in fertility among sequenced men associated with their copy number genotype, but we lack statistical power to detect differences resulting from weak negative selection. We also perform association testing of a diverse set of 341 traits to palindromic copy number without any significant associations. We conclude that large-scale palindrome copy number variation on the Y chromosome has little impact on human phenotype diversity.

Inertia in spatial public goods games under weak selection

Due to limited cognitive skills for perceptual error or other emotional reasons, players may keep their current strategies even if there is a more promising choice. Such behavior inertia has already been studied, but its consequences remained unexplored in the weak selection limit. To fill this gap, we consider a spatial public goods game model where inertia is considered during the imitation process. By using the identity-by-descent method, we present analytical forms of the critical synergy factor $r^\star$, which determines when cooperation is favored. We find that inertia hinders cooperation, which can be explained by the decelerated coarsening process under weak selection. Interestingly, the critical synergy conditions for different updating protocols, including death-birth and birth-death rules, can be formally linked by the extreme limits of the inertia factor. To explore the robustness of our observations, calculations are made for different lattices and group sizes. Monte Carlo simulations also confirm the results.

The conflict between self-interaction and updating passivity in the evolution of cooperation

In social dilemmas under weak selection, the capacity for a player to exhibit updating passivity or interact with its own strategy can lead to conflicting outcomes. The central question is which effect is stronger and how their simultaneous presence influences the evolution of cooperation. We introduce a model that considers both effects using different weight factors. We derive theoretical solutions for the conditions of cooperation success and the cooperation level under weak selection, scanning the complete parameter space. When the weight factors are equally strong, the promoting effect of self-interaction to cooperation surpasses the inhibitory effect of updating passivity. Intriguingly, however, we identify non-monotonous cooperation-supporting effects when the weight of updating passivity increases more rapidly. Our findings are corroborated by Monte Carlo simulations and demonstrate robustness across various game types, including the prisoner’s dilemma, stag-hunt, and snowdrift games.

Fifty shades of greenbeard: robust evolution of altruism based on similarity of complex phenotypes.

We study the evolution of altruistic behaviour under a model where individuals choose to cooperate by comparing a set of continuous phenotype tags. Individuals play a donation game and only donate to other individuals that are sufficiently similar to themselves in a multidimensional phenotype space. We find the generic maintenance of robust altruism when phenotypes are multidimensional. Selection for altruism is driven by the coevolution of individual strategy and phenotype; altruism levels shape the distribution of individuals in phenotype space. Low donation rates induce a phenotype distribution that renders the population vulnerable to the invasion of altruists, whereas high donation rates prime a population for cheater invasion, resulting in cyclic dynamics that maintain substantial levels of altruism. Altruism is therefore robust to invasion by cheaters in the long term in this model. Furthermore, the shape of the phenotype distribution in high phenotypic dimension allows altruists to better resist the invasion by cheaters, and as a result the amount of donation increases with increasing phenotype dimension. We also generalize previous results in the regime of weak selection to two competing strategies in continuous phenotype space, and show that success under weak selection is crucial to success under strong selection in our model. Our results support the viability of a simple similarity-based mechanism for altruism in a well-mixed population.

Evolutionarily rational mutations in structured populations

This paper considers an evolutionary dynamic to study games played in large populations that consist of many identical, potentially small subpopulations, where individuals interact with others in their subpopulation, rarely moving between them.Since long-run behavior depends on the used mutation strategies, a criterion loosely inspired by Maynard Smith's (1972) notion of an evolutionary stable strategy is imposed on mutation strategies – we ask which mutation strategies are robust to “invasions” by others. Non-degenerate mutation strategies lead to behavioral diversity. Populations that use different interaction strategies in different subpopulations can be robust to “invasions,” even if each interaction strategy would not be robust if used by the entire population.Mutation strategies satisfying the proposed criterion always exist. A simple characterization is possible for the case of weak selection. The developed theory yields a rich class of predictions that can be inconsistent with standard concepts like Nash equilibrium. For instance, for finitely repeated games, the model always predicts relatively high payoffs if the stage game is repeated sufficiently often.

Harnessing Radicals in Confined Supramolecular Environments Made Possible by MOFs.

Researching and utilizing radical intermediates in organic synthetic chemistry have innovated discoveries in methodology and theory. Reactions concerning free radical species opened new pathways beyond the frame of the two-electron mechanism while commonly characterized as rampant processes lacking selectivity. As a result, research in this field has always focused on the controllable generation of radical species and determining factors of selectivity. Metal-organic frameworks (MOFs) have emerged as compelling candidates as catalysts in radical chemistry. From a catalytic point of view, the porous nature of MOFs entails an inner phase for the reaction that could offer possibilities for the regulation of reactivity and selectivity. From a material science perspecti ve, MOFs are organic-inorganic hybrid materials that integrate functional units in organic compounds and complex forms in the tunable long-ranged periodic structure. In this account, we summarized our progress in the application of MOFs in radical chemistry in three parts: (1) The generation of radical species; (2) The weak interactions and site selectivity; (3) Regio- and stereo-selectivity. The unique role of MOFs play in these paradigms is demonstrated in a supramolecular narrative through the analyses of the multi-constituent collaboration within the MOF and the interactions between MOFs and the intermediates during the reactions.

Efficient promotion of ethanol complete electrooxidation by anti-poisoning rhodium-bismuth alloy nanodendrites

Alkaline direct ethanol fuel cells are promising future energy technologies with a high energy-density, while the lack of robust anodic electrocatalysts limits their energy conversion efficiency due to serious poisoning and weak selectivity to the C1 pathway of ethanol oxidation reaction (EOR). Herein, rhodium-bismuth alloy nanodendrites (RhBi-NDs) are applied as an efficient and anti-poisoning electrocatalyst for EOR, which exhibit superior electrocatalytic performance, including a mass activity of 1068.9 mA mgRh−1 at 0.71 V, a 31.1% Faradaic efficiency for the C1 product, and an excellent stability of 40 h. The mechanism study indicates that Bi improves the oxidation ability of Rh to toxic intermediates by providing oxygen-containing species. The theoretical calculations disclose the electronic interaction between Bi and Rh, which tunes the adsorption energy of intermediates such as CH3CH2OH*, and CH3CO*, resulting in the suppression of C2 pathway and the promotion of C1 pathway for EOR on RhBi-NDs.

Nontriplet feature of genetic code in Euplotes ciliates is a result of neutral evolution

The triplet nature of the genetic code is considered a universal feature of known organisms. However, frequent stop codons at internal mRNA positions in Euplotes ciliates ultimately specify ribosomal frameshifting by one or two nucleotides depending on the context, thus posing a nontriplet feature of the genetic code of these organisms. Here, we sequenced transcriptomes of eight Euplotes species and assessed evolutionary patterns arising at frameshift sites. We show that frameshift sites are currently accumulating more rapidly by genetic drift than they are removed by weak selection. The time needed to reach the mutational equilibrium is several times longer than the age of Euplotes and is expected to occur after a several-fold increase in the frequency of frameshift sites. This suggests that Euplotes are at an early stage of the spread of frameshifting in expression of their genome. In addition, we find the net fitness burden of frameshift sites to be noncritical for the survival of Euplotes. Our results suggest that fundamental genome-wide changes such as a violation of the triplet character of genetic code can be introduced and maintained solely by neutral evolution.

Novel graphene electrochemical transistors incorporating zirconia inorganic molecular imprinted layer：Design, construction and application for highly sensitive and selective detection of acetaminophen

Owing to their intrinsic amplifying effect together with chemical stability, graphene electrochemical transistor sensors (GECTs) are gaining momentum for sensing applications. However, the surface of GECTs for different detection substances must be modified with different recognition molecules, which was cumbersome and lack a universal method. Molecularly imprinted polymer (MIP) is a kind of polymer with specific recognition function for given molecules. Here, MIP and GECTs were combined to effectively solve the problem of weak selectivity of GECTs, and achieve the high sensitivity and selectivity of MIP-GECTs equipment in detecting acetaminophen (AP) in complex urine environment. A novel molecular imprinting sensor based on Au nanoparticles modified zirconia (ZrO2) inorganic molecular imprinting membrane on reduced graphene oxide (ZrO2-MIP-Au/rGO) was proposed. ZrO2-MIP-Au/rGO was synthesized by one-step electropolymerization using AP as template, ZrO2 precursor as the functional monomer. The –OH group on ZrO2 and the –OH/–CONH– group on AP were easily bonded by hydrogen bonding to form a MIP layer on the surface, which allows the sensor to have a large number of imprinted cavities for AP specific adsorption. As a proof of method, the GECTs based on ZrO2-MIP-Au/rGO functional gate electrode has the characteristics of wide linear range (0.1 nM-4 mM), low detection limit (0.1 nM) and high selectivity for AP detection. These achievements highlight the introduction of specific and selective MIP to GECTs with unique amplification function, which could effectively solve the problem of selectivity of GECTs in complex environments, suggesting the potential of MIP-GECTs in real-time diagnosis.

Weak selection and stochastic evolutionary stability in a stochastic replicator dynamics

It is a very challenging problem whether natural selection is able to effectively resist the continuous disturbance of environmental noise such that the direction or outcome of evolution determined by the deterministic selection pressure will not be changed. By analyzing the impact of weak selection on the evolutionary stability of a stochastic replicator dynamics with n possible pure strategies, we found that the weak selection is able to enhance the evolutionary stability, that is, under weak selection, the stochastic evolutionary stability of the system is determined by the mean payoff matrix. This finding strongly implies that the weak selection should be regarded as an important mechanism to ensure evolutionary stability in stochastic environments.

Shape-selective processing in deep networks: integrating the evidence on perceptual integration

Understanding how deep neural networks resemble or differ from human vision becomes increasingly important with their widespread use in Computer Vision and as models in Neuroscience. A key aspect of human vision is shape: we decompose the visual world into distinct objects, use cues to infer their 3D geometries, and can group several object parts into a coherent whole. Do deep networks use the shape of objects similarly when they classify images? Research on this question has yielded conflicting results, with some studies showing evidence for shape selectivity in deep networks, while others demonstrated clear deficiencies. We argue that these conflicts arise from differences in experimental methods: whether studies use custom images in which only some features are available, images in which different features compete, image pairs that vary along different feature dimensions, or large sets of images to assess how representations vary overall. Each method offers a different, partial view of shape processing. After comparing their advantages and pitfalls, we propose two hypotheses that can reconcile previous results. Firstly, deep networks are sensitive to local, but not global shape. Secondly, the higher layers of deep networks discard some of the shape information that the lower layers are sensitive to. We test these hypotheses by comparing network representations for natural images and silhouettes in which local or global shape is degraded. The results support both hypotheses, but for different networks. Purely feed-forward convolutional networks are unable to integrate shape globally. In contrast, networks with residual or recurrent connections show a weak selectivity for global shape. This motivates further research into recurrent architectures for perceptual integration.

Resurrecting the alternative splicing landscape of archaic hominins using machine learning.

Alternative splicing contributes to adaptation and divergence in many species. However, it has not been possible to directly compare splicing between modern and archaic hominins. Here, we unmask the recent evolution of this previously unobservable regulatory mechanism by applying SpliceAI, a machine-learning algorithm that identifies splice-altering variants (SAVs), to high-coverage genomes from three Neanderthals and a Denisovan. We discover 5,950 putative archaic SAVs, of which 2,186 are archaic-specific and 3,607 also occur in modern humans via introgression (244) or shared ancestry (3,520). Archaic-specific SAVs are enriched in genes that contribute to traits potentially relevant to hominin phenotypic divergence, such as the epidermis, respiration and spinal rigidity. Compared to shared SAVs, archaic-specific SAVs occur in sites under weaker selection and are more common in genes with tissue-specific expression. Further underscoring the importance of negative selection on SAVs, Neanderthal lineages with low effective population sizes are enriched for SAVs compared to Denisovan and shared SAVs. Finally, we find that nearly all introgressed SAVs in humans were shared across the three Neanderthals, suggesting that older SAVs were more tolerated in human genomes. Our results reveal the splicing landscape of archaic hominins and identify potential contributions of splicing to phenotypic differences among hominins.

Evolutionary games and spatial periodicity

Spatial interactions are considered an important factor influencing a variety of evolutionary processes that take place in structured populations. It still remains an open problem to fully understand evolutionary game dynamics on networks except for certain limiting scenarios such as weak selection. Here we study the evolutionary dynamics of spatial games under strong selection where strategy evolution of individuals becomes deterministic in a fashion of winners taking all. We show that the long term behavior of the evolutionary process eventually converges to a particular basin of attraction, which is either a periodic cycle or a single fixed state depending on specific initial conditions and model parameters. In particular, we find that symmetric starting configurations can induce an exceedingly long transient phase encompassing a large number of aesthetic spatial patterns including the prominent kaleidoscopic cooperation. Our finding holds for any population structure and a broad class of finite games beyond the Prisoner’s Dilemma. Our work offers insights into understanding evolutionary dynamics of spatially extended systems ubiquitous in biology and ecology.