Flying Insect Detection and Classification with Inexpensive Sensors

An inexpensive, noninvasive system that could accurately classify flying insects would have important implications for entomological research, and allow for the development of many useful applications in vector and pest control for both medical and agricultural entomology. Given this, the last sixty years have seen many research efforts devoted to this task. To date, however, none of this research has had a lasting impact. In this work, we show that pseudo-acoustic optical sensors can produce superior data; that additional features, both intrinsic and extrinsic to the insect’s flight behavior, can be exploited to improve insect classification; that a Bayesian classification approach allows to efficiently learn classification models that are very robust to over-fitting, and a general classification framework allows to easily incorporate arbitrary number of features. We demonstrate the findings with large-scale experiments that dwarf all previous works combined, as measured by the number of insects and the number of species considered.     

Development of Microplitis croceipes as a biological sensor

Classical conditioning, a form of associative learning, was first described in the vertebrate literature by Pavlov, but has since been documented for a wide variety of insects. Our knowledge of associative learning by insects began with Karl vonFrisch explaining communication among honeybees, Apis mellifera L. (Hymenoptera: Apidae). Since then, the honey bee has provided us with much of what we understand about associative learning in insects and how we relate the theories of learning in vertebrates to insects. Fruit flies, moths, and parasitic wasps are just a few examples of other insects that have been documented with the ability to learn. A novel direction in research on this topic attempts to harness the ability of insects to learn for the development of biological sensors. Parasitic wasps, especially Microplitis croceipes (Cresson) (Hymenoptera: Braconidae), have been conditioned to detect the odors associated with explosives, food toxins, and cadavers. Honeybees and moths have also been associatively conditioned to several volatiles of interest in forensics and national security. In some cases, handheld devices have been developed to harness the insects and observe conditioned behavioral responses to air samples in an attempt to detect target volatiles. Current research on the development of biological sensors with insects is focusing on factors that influence the learning and memory ability of arthropods as well as potential mathematical techniques for improving the interpretation of the behavioral responses to conditioned stimuli. Chemical detection devices using arthropod‐based sensing could be used in situations where trained canines cannot be used (such as toxic environments) or are unavailable, electronic devices are too expensive and/or not of sufficient sensitivity, and when conditioning to target chemicals must be done within minutes of detection. The purpose of this article is to provide a brief review of the development of M. croceipes as a model system for exploring associative learning for the development of biological sensors.     

A novel means of using gene clusters in a two-step empirical Bayes method for predicting classes of samples

MOTIVATION: The classification of samples using gene expression profiles is an important application in areas such as cancer research and environmental health studies. However, the classification is usually based on a small number of samples, and each sample is a long vector of thousands of gene expression levels. An important issue in parametric modeling for so many gene expression levels is the control of the number of nuisance parameters in the model. Large models often lead to intensive or even intractable computation, while small models may be inadequate for complex data.Methodology: We propose a two-step empirical Bayes classification method as a solution to this issue. At the first step, we use the model-based cluster algorithm with a non-traditional purpose of assigning gene expression levels to form abundance groups. At the second step, by assuming the same variance for all the genes in the same group, we substantially reduce the number of nuisance parameters in our statistical model. RESULTS: The proposed model is more parsimonious, which leads to efficient computation under an empirical Bayes estimation procedure. We consider two real examples and simulate data using our method. Desired low classification error rates are obtained even when a large number of genes are pre-selected for class prediction.     

The role of non-fig-wasp insects on fig tree biology,with a proposal of the F phase (Fallen figs)

The two seminal papers by Galil and Eisikowitch describing the development of Ficus flowers and their sycophilous wasps (i.e., phases A-E) have been adopted in several ecological and evolutionary studies on a wide range of fig tree-insect interactions. Their classification, however, is not inclusive enough to encompass all the diversity of insects associated with the fig development, and the impact of this fauna on the fig-fig wasp mutualism is still unexplored. Here we describe the life history of the non-fig-wasp insects and propose an additional phase to fig-development classification, the F phase (Fallen figs). These figs are not consumed by frugivores while still on the parent tree, fall to the ground and turn into a resource for a diverse range of animals. To support the relevance of the F phase, we summarized a 5-years-period of field observations made on different biomes in three continents. Additionally, we compiled data from the literature of non-fig-wasp insects including only insects associated with inflorescences of wild fig tree species. We report 129 species of non-fig-wasp insects feeding on figs; they colonize the figs in different phases of development and some groups rely on the fallen figs to complete their life cycles. Their range of interaction varies from specialists – that use exclusively fig pulp or fig seeds in their diets – to generalists, opportunists and parasitoids species. The formalization of this additional phase will encourage new studies on fig tree ecology and improve our knowledge on the processes that affect the diversification of insects. It will also help us to understand the implications this fauna may have had on the origin and maintenance of mutualistic interactions.     

On the statistical assessment of classifiers using DNA microarray data

Background

In this paper we present a method for the statistical assessment of cancer predictors which make use of gene expression profiles. The methodology is applied to a new data set of microarray gene expression data collected in Casa Sollievo della Sofferenza Hospital, Foggia – Italy. The data set is made up of normal (22) and tumor (25) specimens extracted from 25 patients affected by colon cancer. We propose to give answers to some questions which are relevant for the automatic diagnosis of cancer such as: Is the size of the available data set sufficient to build accurate classifiers? What is the statistical significance of the associated error rates? In what ways can accuracy be considered dependant on the adopted classification scheme? How many genes are correlated with the pathology and how many are sufficient for an accurate colon cancer classification? The method we propose answers these questions whilst avoiding the potential pitfalls hidden in the analysis and interpretation of microarray data.

Results

We estimate the generalization error, evaluated through the Leave-K-Out Cross Validation error, for three different classification schemes by varying the number of training examples and the number of the genes used. The statistical significance of the error rate is measured by using a permutation test. We provide a statistical analysis in terms of the frequencies of the genes involved in the classification. Using the whole set of genes, we found that the Weighted Voting Algorithm (WVA) classifier learns the distinction between normal and tumor specimens with 25 training examples, providing e = 21% (p = 0.045) as an error rate. This remains constant even when the number of examples increases. Moreover, Regularized Least Squares (RLS) and Support Vector Machines (SVM) classifiers can learn with only 15 training examples, with an error rate of e = 19% (p = 0.035) and e = 18% (p = 0.037) respectively. Moreover, the error rate decreases as the training set size increases, reaching its best performances with 35 training examples. In this case, RLS and SVM have error rates of e = 14% (p = 0.027) and e = 11% (p = 0.019). Concerning the number of genes, we found about 6000 genes (p < 0.05) correlated with the pathology, resulting from the signal-to-noise statistic. Moreover the performances of RLS and SVM classifiers do not change when 74% of genes is used. They progressively reduce up to e = 16% (p < 0.05) when only 2 genes are employed. The biological relevance of a set of genes determined by our statistical analysis and the major roles they play in colorectal tumorigenesis is discussed.

Conclusions

The method proposed provides statistically significant answers to precise questions relevant for the diagnosis and prognosis of cancer. We found that, with as few as 15 examples, it is possible to train statistically significant classifiers for colon cancer diagnosis. As for the definition of the number of genes sufficient for a reliable classification of colon cancer, our results suggest that it depends on the accuracy required.     

Nocturnal Homing: Learning Walks in a Wandering Spider?

Homing by the nocturnal Namib Desert spider Leucorchestris arenicola (Araneae: Sparassidae) is comparable to homing in diurnal bees, wasps and ants in terms of path length and layout. The spiders'' homing is based on vision but their basic navigational strategy is unclear. Diurnal homing insects use memorised views of their home in snapshot matching strategies. The insects learn the visual scenery identifying their nest location during learning flights (e.g. bees and wasps) or walks (ants). These learning flights and walks are stereotyped movement patterns clearly different from other movement behaviours. If the visual homing of L. arenicola is also based on an image matching strategy they are likely to exhibit learning walks similar to diurnal insects. To explore this possibility we recorded departures of spiders from a new burrow in an unfamiliar area with infrared cameras and analysed their paths using computer tracking techniques. We found that L. arenicola performs distinct stereotyped movement patterns during the first part of their departures in an unfamiliar area and that they seem to learn the appearance of their home during these movement patterns. We conclude that the spiders perform learning walks and this strongly suggests that L. arenicola uses a visual memory of the burrow location when homing.     

Stochastic models for phylogenetic trees on higher-order taxa

Simple stochastic models for phylogenetic trees on species have been well studied. But much paleontology data concerns time series or trees on higher-order taxa, and any broad picture of relationships between extant groups requires use of higher-order taxa. A coherent model for trees on (say) genera should involve both a species-level model and a model for the classification scheme by which species are assigned to genera. We present a general framework for such models, and describe three alternate classification schemes. Combining with the species-level model of Aldous and Popovic (Adv Appl Probab 37:1094–1115, 2005), one gets models for higher-order trees, and we initiate analytic study of such models. In particular we derive formulas for the lifetime of genera, for the distribution of number of species per genus, and for the offspring structure of the tree on genera. David Aldous’s research was supported by NSF Grant DMS-0704159.     

Deep Learning for Population Genetic Inference

Given genomic variation data from multiple individuals, computing the likelihood of complex population genetic models is often infeasible. To circumvent this problem, we introduce a novel likelihood-free inference framework by applying deep learning, a powerful modern technique in machine learning. Deep learning makes use of multilayer neural networks to learn a feature-based function from the input (e.g., hundreds of correlated summary statistics of data) to the output (e.g., population genetic parameters of interest). We demonstrate that deep learning can be effectively employed for population genetic inference and learning informative features of data. As a concrete application, we focus on the challenging problem of jointly inferring natural selection and demography (in the form of a population size change history). Our method is able to separate the global nature of demography from the local nature of selection, without sequential steps for these two factors. Studying demography and selection jointly is motivated by Drosophila, where pervasive selection confounds demographic analysis. We apply our method to 197 African Drosophila melanogaster genomes from Zambia to infer both their overall demography, and regions of their genome under selection. We find many regions of the genome that have experienced hard sweeps, and fewer under selection on standing variation (soft sweep) or balancing selection. Interestingly, we find that soft sweeps and balancing selection occur more frequently closer to the centromere of each chromosome. In addition, our demographic inference suggests that previously estimated bottlenecks for African Drosophila melanogaster are too extreme.     

Phylogenetic relationships of phrynosomatid lizards based on nuclear and mitochondrial data,and a revised phylogeny for Sceloporus

Phrynosomatid lizards are among the most common and diverse groups of reptiles in western North America, Mexico, and Central America. Phrynosomatidae includes 136 species in 10 genera. Phrynosomatids are used as model systems in many research programs in evolution and ecology, and much of this research has been undertaken in a comparative phylogenetic framework. However, relationships among many phrynosomatid genera are poorly supported and in conflict between recent studies. Further, previous studies based on mitochondrial DNA sequences suggested that the most species-rich genus (Sceloporus) is possibly paraphyletic with respect to as many as four other genera (Petrosaurus, Sator, Urosaurus, and Uta). Here, we collect new sequence data from five nuclear genes and combine them with published data from one additional nuclear gene and five mitochondrial gene regions. We compare trees from nuclear and mitochondrial data from 37 phrynosomatid taxa, including a “species tree” (from BEST) for the nuclear data. We also present a phylogeny for 122 phrynosomatid species based on maximum likelihood analysis of the combined data, which provides a strongly-supported hypothesis for relationships among most phrynosomatid genera and includes most phrynosomatid species. Our results strongly support the monophyly of Sceloporus (including Sator) and many of the relationships within it. We present a new classification for phrynosomatid lizards and the genus Sceloporus, and offer a new tree with branch lengths for use in comparative studies.     

The effects of juvenile hormone on Lasius niger reproduction

Reproduction has been shown to be costly for survival in a wide diversity of taxa. The resulting trade-off, termed the reproduction-survival trade-off, is thought to be one of the most fundamental forces of life-history evolution. In insects the pleiotropic effect of juvenile hormone (JH), antagonistically regulating reproduction and pathogen resistance, is suggested to underlie this phenomenon. In contrast to the majority of insects, reproductive individuals in many eusocial insects defy this trade-off and live both long and prosper. By remodelling the gonadotropic effects of JH in reproductive regulation, the queens of the long-lived black garden ant Lasius niger (living up to 27 years), have circumvented the reproduction-survival trade off enabling them to maximize both reproduction and pathogen resistance simultaneously. In this study we measure fertility, vitellogenin gene expression and protein levels after experimental manipulation of hormone levels. We use these measurements to investigate the mechanistic basis of endocrinological role remodelling in reproduction and determine how JH suppresses reproduction in this species, rather then stimulating it, like in the majority of insects. We find that JH likely inhibits three key aspects of reproduction both during vitellogenesis and oogenesis, including two previously unknown mechanisms. In addition, we document that juvenile hormone, as in the majority of insects, has retained some stimulatory function in regulating vitellogenin expression. We discuss the evolutionary consequences of this complex regulatory architecture of reproduction in L. niger, which might enable the evolution of similar reproductive phenotypes by alternate regulatory pathways, and the surprising flexibility regulatory role of juvenile hormone in this process.     

The value of protein structure classification information—Surveying the scientific literature

The Structural Classification of Proteins (SCOP) and Class, Architecture, Topology, Homology (CATH) databases have been valuable resources for protein structure classification for over 20 years. Development of SCOP (version 1) concluded in June 2009 with SCOP 1.75. The SCOPe (SCOP–extended) database offers continued development of the classic SCOP hierarchy, adding over 33,000 structures. We have attempted to assess the impact of these two decade old resources and guide future development. To this end, we surveyed recent articles to learn how structure classification data are used. Of 571 articles published in 2012–2013 that cite SCOP, 439 actually use data from the resource. We found that the type of use was fairly evenly distributed among four top categories: A) study protein structure or evolution (27% of articles), B) train and/or benchmark algorithms (28% of articles), C) augment non‐SCOP datasets with SCOP classification (21% of articles), and D) examine the classification of one protein/a small set of proteins (22% of articles). Most articles described computational research, although 11% described purely experimental research, and a further 9% included both. We examined how CATH and SCOP were used in 158 articles that cited both databases: while some studies used only one dataset, the majority used data from both resources. Protein structure classification remains highly relevant for a diverse range of problems and settings. Proteins 2015; 83:2025–2038. © 2015 The Authors. Proteins: Structure, Function, and Bioinformatics Published by Wiley Periodicals, Inc.     

AUTOMATIC ACOUSTIC IDENTIFICATION OF SINGING INSECTS

ABSTRACT

We report on our research efforts towards developing efficient equipment for the automatic recognition of insects using only the acoustic modality. Specifically, we deal with three groups of insects, namely the crickets, cicadas and katydids. Inspired by well-documented tactics of speech processing, the signal processing employed in the present work is elaborated further with respect to the sound production mechanisms of insects. In order to improve the practical efficacy of our equipment, we adopt a score-level fusion of classifiers with non-parametric (probabilistic neural network) and parametric (Gaussian mixture models) estimation of the probability density function. An efficient hierarchic classification scheme is introduced, where the identification of unlabelled input takes place at various levels of hierarchy, such as suborder, family, subfamily, genus and species. We evaluate the practical significance of our approach on a large and well-documented catalogue of recordings of crickets, cicadas and katydids. For the hierarchic classification scheme, we report identification accuracy that exceeds 99% at suborder and family levels. In the straight classification scheme, we report accuracy of 90% for 307 species.     

DCMD: Distance-based classification using mixture distributions on microbiome data

Current advances in next-generation sequencing techniques have allowed researchers to conduct comprehensive research on the microbiome and human diseases, with recent studies identifying associations between the human microbiome and health outcomes for a number of chronic conditions. However, microbiome data structure, characterized by sparsity and skewness, presents challenges to building effective classifiers. To address this, we present an innovative approach for distance-based classification using mixture distributions (DCMD). The method aims to improve classification performance using microbiome community data, where the predictors are composed of sparse and heterogeneous count data. This approach models the inherent uncertainty in sparse counts by estimating a mixture distribution for the sample data and representing each observation as a distribution, conditional on observed counts and the estimated mixture, which are then used as inputs for distance-based classification. The method is implemented into a k-means classification and k-nearest neighbours framework. We develop two distance metrics that produce optimal results. The performance of the model is assessed using simulated and human microbiome study data, with results compared against a number of existing machine learning and distance-based classification approaches. The proposed method is competitive when compared to the other machine learning approaches, and shows a clear improvement over commonly used distance-based classifiers, underscoring the importance of modelling sparsity for achieving optimal results. The range of applicability and robustness make the proposed method a viable alternative for classification using sparse microbiome count data. The source code is available at https://github.com/kshestop/DCMD for academic use.     

Passive acoustic monitoring of animal populations with transfer learning

Progress in deep learning, more specifically in using convolutional neural networks (CNNs) for the creation of classification models, has been tremendous in recent years. Within bioacoustics research, there has been a large number of recent studies that use CNNs. Designing CNN architectures from scratch is non-trivial and requires knowledge of machine learning. Furthermore, hyper-parameter tuning associated with CNNs is extremely time consuming and requires expensive hardware. In this paper we assess whether it is possible to build good bioacoustic classifiers by adapting and re-using existing CNNs pre-trained on the ImageNet dataset – instead of designing them from scratch, a strategy known as transfer learning that has proved highly successful in other domains. This study is a first attempt to conduct a large-scale investigation on how transfer learning can be used for passive acoustic monitoring (PAM), to simplify the implementation of CNNs and the design decisions when creating them, and to remove time consuming hyper-parameter tuning phases. We compare 12 modern CNN architectures across 4 passive acoustic datasets that target calls of the Hainan gibbon Nomascus hainanus, the critically endangered black-and-white ruffed lemur Varecia variegata, the vulnerable Thyolo alethe Chamaetylas choloensis, and the Pin-tailed whydah Vidua macroura. We focus our work on data scarcity issues by training PAM binary classification models very small datasets, with as few as 25 verified examples. Our findings reveal that transfer learning can result in up to 82% F1 score while keeping CNN implementation details to a minimum, thus rendering this approach accessible, easier to design, and speeding up further vocalisation annotations to create PAM robust models.     

Estimating animal population density using passive acoustics

Reliable estimation of the size or density of wild animal populations is very important for effective wildlife management, conservation and ecology. Currently, the most widely used methods for obtaining such estimates involve either sighting animals from transect lines or some form of capture‐recapture on marked or uniquely identifiable individuals. However, many species are difficult to sight, and cannot be easily marked or recaptured. Some of these species produce readily identifiable sounds, providing an opportunity to use passive acoustic data to estimate animal density. In addition, even for species for which other visually based methods are feasible, passive acoustic methods offer the potential for greater detection ranges in some environments (e.g. underwater or in dense forest), and hence potentially better precision. Automated data collection means that surveys can take place at times and in places where it would be too expensive or dangerous to send human observers. Here, we present an overview of animal density estimation using passive acoustic data, a relatively new and fast‐developing field. We review the types of data and methodological approaches currently available to researchers and we provide a framework for acoustics‐based density estimation, illustrated with examples from real‐world case studies. We mention moving sensor platforms (e.g. towed acoustics), but then focus on methods involving sensors at fixed locations, particularly hydrophones to survey marine mammals, as acoustic‐based density estimation research to date has been concentrated in this area. Primary among these are methods based on distance sampling and spatially explicit capture‐recapture. The methods are also applicable to other aquatic and terrestrial sound‐producing taxa. We conclude that, despite being in its infancy, density estimation based on passive acoustic data likely will become an important method for surveying a number of diverse taxa, such as sea mammals, fish, birds, amphibians, and insects, especially in situations where inferences are required over long periods of time. There is considerable work ahead, with several potentially fruitful research areas, including the development of (i) hardware and software for data acquisition, (ii) efficient, calibrated, automated detection and classification systems, and (iii) statistical approaches optimized for this application. Further, survey design will need to be developed, and research is needed on the acoustic behaviour of target species. Fundamental research on vocalization rates and group sizes, and the relation between these and other factors such as season or behaviour state, is critical. Evaluation of the methods under known density scenarios will be important for empirically validating the approaches presented here.     

Generating functional protein variants with variational autoencoders

The vast expansion of protein sequence databases provides an opportunity for new protein design approaches which seek to learn the sequence-function relationship directly from natural sequence variation. Deep generative models trained on protein sequence data have been shown to learn biologically meaningful representations helpful for a variety of downstream tasks, but their potential for direct use in the design of novel proteins remains largely unexplored. Here we show that variational autoencoders trained on a dataset of almost 70000 luciferase-like oxidoreductases can be used to generate novel, functional variants of the luxA bacterial luciferase. We propose separate VAE models to work with aligned sequence input (MSA VAE) and raw sequence input (AR-VAE), and offer evidence that while both are able to reproduce patterns of amino acid usage characteristic of the family, the MSA VAE is better able to capture long-distance dependencies reflecting the influence of 3D structure. To confirm the practical utility of the models, we used them to generate variants of luxA whose luminescence activity was validated experimentally. We further showed that conditional variants of both models could be used to increase the solubility of luxA without disrupting function. Altogether 6/12 of the variants generated using the unconditional AR-VAE and 9/11 generated using the unconditional MSA VAE retained measurable luminescence, together with all 23 of the less distant variants generated by conditional versions of the models; the most distant functional variant contained 35 differences relative to the nearest training set sequence. These results demonstrate the feasibility of using deep generative models to explore the space of possible protein sequences and generate useful variants, providing a method complementary to rational design and directed evolution approaches.     

An exception to the rule: common vampire bats do not learn taste aversions

Conditioned taste aversions function by preventing an organism from ingesting a food previously associated with gastrointestinal malaise. Taste-aversion learning has been observed in many animals: molluscs to mammals, insects to birds. However, among mammals, neither bats nor monophagous species have been investigated adequately. Here we show that although three dietary generalists (one insectivorous and two frugivorous bats) readily acquired taste aversions, the common vampire bat, Desmodus rotundus, a monophageous feeder on vertebrate blood, did not learn to associate a novel flavour with aversive gastrointestinal events. We interpret these data as consistent with the hypothesis that taste aversions are an adaptive specialization of learning. Copyright 2003 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.      

Demographic consequences of inflorescence-feeding insects for Liatris cylindracea, an iteroparous perennial

While floral herbivores and predispersal seed predators often reduce plant reproductive output, their role in limiting plant fitness and population growth is less clear, especially for iteroparous perennial plant species. In this study we experimentally excluded floral herbivores and predispersal seed predators (insecticide spray versus water control) over a 2-year period to examine the effect of inflorescence-feeding insects on levels of seed production, seedling emergence, and juvenile establishment for Liatris cylindracea, an iteroparous perennial plant. In addition, we collected detailed demographic data on all life stage transitions for an additional set of individuals in the same population over 4 years. We used the experimental and demographic data to construct stochastic individual-based simulations to evaluate the overall effect of inflorescence-feeding insects on adult recruitment per maternal plant (a fitness component) and population growth rate. The insect exclusion experiments showed that damage due to insects decreased seed production, seedling emergence, and juvenile establishment for both years' experiments. These results indicate that recruitment was seed-limited through juvenile establishment, and that inflorescence-feeding insects influenced the degree of seed limitation. Results of the individual-based simulation models, which included individual demographic and temporal stochasticity, showed that inflorescence-feeding insects negatively affected the number of adult offspring per maternal plant recruited into the population and population growth rate for both years' experiments. Taken together, the results of the experimental exclusions and the individual-based models indicate that inflorescence-feeding insects can influence population growth rate, and have the potential to act as a selective force for the evolution of traits in this plant species.     

Learning brain dynamics for decoding and predicting individual differences

Insights from functional Magnetic Resonance Imaging (fMRI), as well as recordings of large numbers of neurons, reveal that many cognitive, emotional, and motor functions depend on the multivariate interactions of brain signals. To decode brain dynamics, we propose an architecture based on recurrent neural networks to uncover distributed spatiotemporal signatures. We demonstrate the potential of the approach using human fMRI data during movie-watching data and a continuous experimental paradigm. The model was able to learn spatiotemporal patterns that supported 15-way movie-clip classification (∼90%) at the level of brain regions, and binary classification of experimental conditions (∼60%) at the level of voxels. The model was also able to learn individual differences in measures of fluid intelligence and verbal IQ at levels comparable to that of existing techniques. We propose a dimensionality reduction approach that uncovers low-dimensional trajectories and captures essential informational (i.e., classification related) properties of brain dynamics. Finally, saliency maps and lesion analysis were employed to characterize brain-region/voxel importance, and uncovered how dynamic but consistent changes in fMRI activation influenced decoding performance. When applied at the level of voxels, our framework implements a dynamic version of multivariate pattern analysis. Our approach provides a framework for visualizing, analyzing, and discovering dynamic spatially distributed brain representations during naturalistic conditions.