Extending nonlinear analysis to short ecological time series

Nonlinearity is important and ubiquitous in ecology. Though detectable in principle, nonlinear behavior is often difficult to characterize, analyze, and incorporate mechanistically into models of ecosystem function. One obvious reason is that quantitative nonlinear analysis tools are data intensive (require long time series), and time series in ecology are generally short. Here we demonstrate a useful method that circumvents data limitation and reduces sampling error by combining ecologically similar multispecies time series into one long time series. With this technique, individual ecological time series containing as few as 20 data points can be mined for such important information as (1) significantly improved forecast ability, (2) the presence and location of nonlinearity, and (3) the effective dimensionality (the number of relevant variables) of an ecological system.     

Artificial neural network learning of nonstationary behavior in time series

We refine and complement a previously-proposed artificial neural network method for learning hidden signals forcing nonstationary behavior in time series. The method adds an extra input unit to the network and feeds it with the proposed profile for the unknown perturbing signal. The correct time evolution of this new input parameter is learned simultaneously with the intrinsic stationary dynamics underlying the series, which is accomplished by minimizing a suitably-defined error function for the training process. We incorporate here the use of validation data, held out from the training set, to accurately determine the optimal value of a hyperparameter required by the method. Furthermore, we evaluate this algorithm in a controlled situation and show that it outperforms other existing methods in the literature. Finally, we discuss a preliminary application to the real-world sunspot time series and link the obtained hidden perturbing signal to the secular evolution of the solar magnetic field.     

Symbolic dynamics for identifying similarity between rhythms of ecological time series

Explaining the associations between animal populations or between population and environmental signals is an important challenge. The time series that quantify animal populations are often complex, nonlinear, noisy and non‐stationary. These characteristics may make it inappropriate to use traditional techniques when analysing these time series and their mutual dependencies. Here I propose to use symbolic dynamics and techniques from Information Theory to evaluate the degree of dynamic cohesion between time series fluctuations. The main idea is to check whether two (or more) signals tend to oscillate simultaneously, rising and falling together with the same rhythm. Based on synthetic and real time series, I demonstrate that this method is robust to the presence of noise and to the short length of the analysed time series and gives relevant information about the weak relationships between different series. Furthermore, this method appears as simple as classical cross‐correlation and outperforms it in the analysed examples.     

Conditioned reflex analysis of functional connections between the hippocampus and limbic system structures

Instrumental defensive conditioned reflex elaborated in dogs to stimulation of the dorsal and ventral hippocampus, also appeared, due to generalization, in response to stimulation of a number of limbic structures. Two types of changes in the generalization effects (estimated by parameters of motor conditioned reaction) were observed in the course of conditioned reflex stabilization: enhancement (in response to stimulation of the lateral hypothalamus, lateral nucleus of septum, limbic cortex) and weakening toward complete disappearance (in response to stimulation of the medial nucleus of amygdala and medial hypothalamus). Manifestation of the generalization phenomenon from the brain structures, which are not involved initially into conditioned activity, suggests the existence of close functional connections between these structures and the hippocampus.     

Modelling algal growth and transport in rivers: a comparison of time series analysis,dynamic mass balance and neural network techniques

Algae present considerable problems for river qualitymanagers and water suppliers and methods to predicttheir behaviour, growth and transport can assist inoperational management. Alternative techniques existfor predicting algal response and three approacheshave been compared and applied to data from six sitesalong the River Thames. These techniques include timeseries analysis, dynamic mass balance and growthequations and neural network approaches. It is shownthat neural network techniques offer a new approachrequiring less intuitive knowledge but predictivecapability is not improved greatly compared to otherapproaches. Neural networks enable models to bedeveloped along all six reaches of the RiverThames.     

Functional connections between cells as revealed by dye-coupling with a highly fluorescent naphthalimide tracer

This report describes a method of marking nerve cells which is approximately 100 times more sensitive than those previously available. The method depends upon intracellular injection of a new, highly fluorescent dye, Lucifer Yellow CH, which can be viewed both in living tissue and after fixation and embedding. The intense fluorescence of the dye makes injected neurons visible in cleared wholemounts, where the complex three-dimensional structure of neurons is readily apparent.Three new observations have been made with Lucifer Yellow. First, many of the invertebrate neurons studied possess an extensive and complex array of fine processes not visible with other techniques. Second, dye spreads rapidly within an injected cell. Third, dye frequently spreads from the injected cell directly to certain other cells. The movement of dye from cell to cell, termed “dye-coupling,” occurred primarily, but not exclusively, between cells known to be electrically coupled.Dye-coupling in the turtle retina revealed striking and distinctive patterns of connections. Type I horizontal cells appear to be multiply connected to each other in an extensive net. Type II horizontal cells are often connected to each other in a hexagonal array. Individual type I and type II cells, widely separated, are frequently dye-coupled; in one case, they were connected by a dyefilled axon.Dye-coupling, readily observed because of the low molecular weight and the intense fluorescence of the new dye, may serve as a general method of tracing certain functional connections by morphological means, and of studying the transfer of small molecules between cells. Preliminary results suggest that systems of dye-coupled cells are substantially more common than was previously believed.     

Nonantagonistic interactions between the sexes revealed by the ecological consequences of reproductive traits

In addition to the obvious role reproductive traits play in mating-system evolution, reproductive characters can also have critical ecological or life history consequences. In this study we examine the ecological consequences of mating for female cactophilic Drosophila to test different hypotheses about the processes driving divergence in reproductive characters. Comparisons between intra- and interpopulation matings suggest that population differences in mating benefits, namely increased desiccation resistance in mated females, is not solely attributable to either a male or female-specific reproductive trait. Instead, the results indicate that increased desiccation resistance is a product of a male-female postmating-prezygotic interactions. The results underscore that postmating-prezygotic interactions can serve as an arena for the evolution of male characters that confer substantial benefits to females, not just costs arising from sexual conflict. Variation in the relative benefits conferred by mating between intra- and interpopulation matings also suggests that the relationship between speciation and divergence in reproductive characters via male-female interaction will be difficult to predict.     

On-line estimation of the time constant of oxygen electrodes by time series analysis

Summary The time constant of an oxygen electrode was estimated on-line by fitting a first order autoregressive model to the normal process noise. The method allows the time constant to be determined under normal operating conditions, thus the effect of the outside liquid film can be accounted for. This has applications to the determination of K_La.     

Conserved antigenic sites between MERS-CoV and Bat-coronavirus are revealed through sequence analysis

Background

MERS-CoV is a newly emerged human coronavirus reported closely related with HKU4 and HKU5 Bat coronaviruses. Bat and MERS corona-viruses are structurally related. Therefore, it is of interest to estimate the degree of conserved antigenic sites among them. It is of importance to elucidate the shared antigenic-sites and extent of conservation between them to understand the evolutionary dynamics of MERS-CoV.

Results

Multiple sequence alignment of the spike (S), membrane (M), enveloped (E) and nucleocapsid (N) proteins was employed to identify the sequence conservation among MERS and Bat (HKU4, HKU5) coronaviruses. We used various in silico tools to predict the conserved antigenic sites. We found that MERS-CoV shared 30 % of its S protein antigenic sites with HKU4 and 70 % with HKU5 bat-CoV. Whereas 100 % of its E, M and N protein’s antigenic sites are found to be conserved with those in HKU4 and HKU5.

Conclusion

This sharing suggests that in case of pathogenicity MERS-CoV is more closely related to HKU5 bat-CoV than HKU4 bat-CoV. The conserved epitopes indicates their evolutionary relationship and ancestry of pathogenicity.

     

Cluster analysis of respiratory time series

We have investigated the respiratory control system with the hypothesis that, although many variables such as minute ventilation (V _I), tidal volume (V _T),breathing period (T _T),inspiratory duration (T _I),and exspiratory duration (T _E),may be observed, the controller functions more simply by manipulating only 2 or 3 of these. Thus, if tidal volume is the only independent variable, T _Ibeing determined by the off-switch threshold, these variables should have very similar time courses. Anesthetized dogs were subjected to CO₂ breathing and carotid sinus perfusion to stimulate both chemoreceptors. The time series of the variables V _I, V _T, T_T, T_Eand T _Ias well as P _{A CO 2}were determined on a breath by breath basis. Derived characteristics of these time series were compared using Cluster Analysis and the latent dimensionality of respiratory control determined by Factor Analysis. The characteristics of the time series clustered into 4 groups: magnitude (of the response), speed, variability and relative change. One respiratory factor accounted for 86% of the variance for the variability characteristics, 2 factors for magnitude (91%) and relative change (85%) and 3 factors for speed (89%). The respiratory variables were analysed for each of the 4 groups of characteristics with the following results: V _Tand T _I clustered together only for the magnitude and relative change characteristics where as T _Tand T _Eclustered closely for all four characteristics. One latent factor was associated with the [T _T-T_E]group and the other usually with P _{A CO 2}.Supported by USPHS 5t01 01919-05, NIH HL 12564 and GM 07033     

Short-term prediction of chaotic time series by using RBF network with regression weights

We propose a framework for constructing and training a radial basis function (RBF) neural network. The structure of the gaussian functions is modified using a pseudo-gaussian function (PG) in which two scaling parameters sigma are introduced, which eliminates the symmetry restriction and provides the neurons in the hidden layer with greater flexibility with respect to function approximation. We propose a modified PG-BF (pseudo-gaussian basis function) network in which the regression weights are used to replace the constant weights in the output layer. For this purpose, a sequential learning algorithm is presented to adapt the structure of the network, in which it is possible to create a new hidden unit and also to detect and remove inactive units. A salient feature of the network systems is that the method used for calculating the overall output is the weighted average of the output associated with each receptive field. The superior performance of the proposed PG-BF system over the standard RBF are illustrated using the problem of short-term prediction of chaotic time series.     

Electrophysiological analysis of the functional organization of cortico-cerebellar connections

The effect of stimulation of cortical association (orbito-frontal, parietal) and projection (auditory, sensomotor) areas on the activity of Purkinje neurons of the cerebellar cortex was studied in adult cats anesthetized with pentobarbital, with or without chloralose. These responses were compared with those to peripheral stimuli. Definite similarity was found between the responses of Purkinje cells to different cortical (association and projection) stimuli as regards both the types of responses of the neurons and their ability to respond. No similarity was observed in the responses of Purkinje cells to peripheral (visual, auditory, electrodermal) stimulation. Whereas almost identical numbers of neurons (over 50%) were excited in response to the different forms of cortical stimulation, the ability of the neurons to respond to peripheral stimuli differed considerably: 44.6% of neurons responded to electrodermal stimulation, 34.2% to auditory, and 18.8% to visual.Medical Institute, Kemerovo. Translated from Neirofiziologiya, Vol. 8, No. 5, pp. 483–489, September–October, 1976.     

Probabilities of spurious connections in gene networks: application to expression time series

MOTIVATION: The reconstruction of gene networks from gene-expression microarrays is gaining popularity as methods improve and as more data become available. The reliability of such networks could be judged by the probability that a connection between genes is spurious, resulting from chance fluctuations rather than from a true biological relationship. RESULTS: Unlike the false discovery rate and positive false discovery rate, the decisive false discovery rate (dFDR) is exactly equal to a conditional probability without assuming independence or the randomness of hypothesis truth values. This property is useful not only in the common application to the detection of differential gene expression, but also in determining the probability of a spurious connection in a reconstructed gene network. Estimators of the dFDR can estimate each of three probabilities: (1) The probability that two genes that appear to be associated with each other lack such association. (2) The probability that a time ordering observed for two associated genes is misleading. (3) The probability that a time ordering observed for two genes is misleading, either because they are not associated or because they are associated without a lag in time. The first probability applies to both static and dynamic gene networks, and the other two only apply to dynamic gene networks.     

Metabolic network analysis revealed distinct routes of deletion effects between essential and non-essential genes

Interest in essential genes has arisen recently given their importance in antimicrobial drug development. Although knockouts of essential genes are commonly known to cause lethal phenotypes, there is insufficient understanding on the intermediate changes followed by genetic perturbation and to what extent essential genes correlate to other genes. Here, we characterized the gene knockout effects by using a list of affected genes, termed as 'damage lists'. These damage lists were identified through a refined cascading failure approach that was based on a previous topological flux balance analysis. Using an Escherichia coli metabolic network, we incorporated essentiality information into damage lists and revealed that the knockout of an essential gene mainly affects a large range of other essential genes whereas knockout of a non-essential gene only interrupts other non-essential genes. Also, genes sharing common damage lists tend to have the same essentiality. We extracted 72 core functional modules from the common damage lists of essential genes and demonstrated their ability to halt essential metabolites production. Overall, our network analysis revealed that essential and non-essential genes propagated their deletion effects via distinct routes, conferring mechanistic explanation to the observed lethality phenotypes of essential genes.     

Central connections of Limulus ventral photoreceptors revealed by intracellular staining

We stained the central terminations of Limulus ventral photoreceptors by intracellular injection of cobalt chloride into the cell bodies. Axons of these photoreceptors enter the protocerebrum via the ventral optic nerve and pass to the medulla. As they reach the surface of the medullar neuropil they branch profusely in fine processes with intermittent varicosities. Each axonal arborization covers about 0.01-0.02 mm2 of this surface immediately adjacent to the medullar ganglion cell layer. Each point on the surface of the medullar neuropil receives, on the average, input from about 6 ventral photoreceptor axons.     

Multiple etiologies for Alzheimer disease are revealed by segregation analysis.

We have evaluated several transmission models for Alzheimer disease (AD), using the logistic regressive approach in 401 nuclear families of consecutively ascertained and rigorously diagnosed probands. Models postulating no major gene effect, random environmental transmission, recessive inheritance, and sporadic occurrence were rejected under varied assumptions regarding the associations among sex, age, and major gene susceptibility. Transmission of the disorder was not fully explained by a single Mendelian model for all families. Stratification of families as early- and late-onset by using the median of family mean onset ages showed that, regardless of the model studied, two groups of families fit better than a single group. AD in early-onset families is transmitted as an autosomal dominant trait with full penetrance in both sexes and has a gene frequency of 1.5%. Dominant inheritance also gave the best fit of the data in late-onset families, but this hypothesis was rejected, suggesting the presence of heterogeneity within this subset. Our study also revealed that genetically nonsusceptible males and females develop AD, indicating the presence of phenocopies within early-onset and late-onset groups. Moreover, our results suggest that the higher risk to females is not solely due to their increased longevity.     

A scale of functional divergence for yeast duplicated genes revealed from analysis of the protein-protein interaction network

Background  

Studying the evolution of the function of duplicated genes usually implies an estimation of the extent of functional conservation/divergence between duplicates from comparison of actual sequences. This only reveals the possible molecular function of genes without taking into account their cellular function(s). We took into consideration this latter dimension of gene function to approach the functional evolution of duplicated genes by analyzing the protein-protein interaction network in which their products are involved. For this, we derived a functional classification of the proteins using PRODISTIN, a bioinformatics method allowing comparison of protein function. Our work focused on the duplicated yeast genes, remnants of an ancient whole-genome duplication.