Search

In this paper, we propose a framework (ALMO) for any-shot learning from multi-modal observations. Using training data containing both objects (inputs) and class attributes (side information) from multiple modalities, ALMO embeds the high-dimensional data into a common stochastic latent space using modality-specific encoders. Subsequently, a non-parametric classifier is trained to predict the class labels of the objects. We perform probabilistic data fusion to combine the modalities in the stochastic latent space and learn class conditional distributions for improved generalization and scalability. We formulate ALMO for both few-shot and zero-shot classification tasks, demonstrating significant improvement in recognition performance on the Omniglot and CUB-200 datasets as compared to state-of-the-art baselines.

Abstract Using data from a longitudinal viral challenge study, we find that the post-exposure viral shedding and symptom severity are associated with a novel measure of pre-exposure cognitive performance variability (CPV), defined before viral exposure occurs. Each individual’s CPV score is computed from data collected from a repeated NeuroCognitive Performance Test (NCPT) over a 3 day pre-exposure period. Of the 18 NCPT measures reported by the tests, 6 contribute materially to the CPV score, prospectively differentiating the high from the low shedders. Among these 6 are the 4 clinical measures digSym-time, digSym-correct, trail-time, and reaction-time, commonly used for assessing cognitive executive functioning. CPV is found to be correlated with stress and also with several genes previously reported to be associated with cognitive development and dysfunction. A perturbation study over the number and timing of NCPT sessions indicates that as few as 5 sessions is sufficient to maintain high association between the CPV score and viral shedding, as long as the timing of these sessions is balanced over the three pre-exposure days. Our results suggest that variations in cognitive function are closely related to immunity and susceptibility to severe infection. Further studying these relationships may help us better understand the links between neurocognitive and neuroimmune systems which is timely in this COVID-19 pandemic era.

Abstract Background Development of new methods for analysis of protein–protein interactions (PPIs) at molecular and nanometer scales gives insights into intracellular signaling pathways and will improve understanding of protein functions, as well as other nanoscale structures of biological and abiological origins. Recent advances in computational tools, particularly the ones involving modern deep learning algorithms, have been shown to complement experimental approaches for describing and rationalizing PPIs. However, most of the existing works on PPI predictions use protein-sequence information, and thus have difficulties in accounting for the three-dimensional organization of the protein chains. Results In this study, we address this problem and describe a PPI analysis based on a graph attention network, named Struct2Graph, for identifying PPIs directly from the structural data of folded protein globules. Our method is capable of predicting the PPI with an accuracy of 98.89% on the balanced set consisting of an equal number of positive and negative pairs. On the unbalanced set with the ratio of 1:10 between positive and negative pairs, Struct2Graph achieves a fivefold cross validation average accuracy of 99.42%. Moreover, Struct2Graph can potentially identify residues that likely contribute to the formation of the protein–protein complex. The identification of important residues is tested for two different interaction types: (a) Proteins with multiple ligands competing for the same binding area, (b) Dynamic protein–protein adhesion interaction. Struct2Graph identifies interacting residues with 30% sensitivity, 89% specificity, and 87% accuracy. Conclusions In this manuscript, we address the problem of prediction of PPIs using a first of its kind, 3D-structure-based graph attention network (code available at https://github.com/baranwa2/Struct2Graph ). Furthermore, the novel mutual attention mechanism provides insights into likely interaction sites through its unsupervised knowledge selection process. This study demonstrates that a relatively low-dimensional feature embedding learned from graph structures of individual proteins outperforms other modern machine learning classifiers based on global protein features. In addition, through the analysis of single amino acid variations, the attention mechanism shows preference for disease-causing residue variations over benign polymorphisms, demonstrating that it is not limited to interface residues.

Adversarial attacks against deep neural networks (DNNs) are continuously evolving, requiring increasingly powerful defense strategies. We develop a novel adversarial defense framework inspired by the adaptive immune system: the Robust Adversarial Immune-inspired Learning System (RAILS). Initializing a population of exemplars that is balanced across classes, RAILS starts from a uniform label distribution that encourages diversity and uses an evolutionary optimization process to adaptively adjust the predictive label distribution in a manner that emulates the way the natural immune system recognizes novel pathogens. RAILS’ evolutionary optimization process explicitly captures the tradeoff between robustness (diversity) and accuracy (specificity) of the network, and represents a new immune-inspired perspective on adversarial learning. The benefits of RAILS are empirically demonstrated under eight types of adversarial attacks on a DNN adversarial image classifier for several benchmark datasets, including: MNIST; SVHN; CIFAR-10; and CIFAR-10. We find that PGD is the most damaging attack strategy and that for this attack RAILS is significantly more robust than other methods, achieving improvements in adversarial robustness by $\geq 5.62\%, 12.5\%$ , 10.32%, and 8.39%, on these respective datasets, without appreciable loss of classification accuracy. Codes for the results in this paper are available at https://github.com/wangren09/RAILS.

K-Nearest Neighbor (kNN)-based deep learning methods have been applied to many applications due to their simplicity and geometric interpretability. However, the robustness of kNN-based deep classification models has not been thoroughly explored and kNN attack strategies are underdeveloped. In this paper, we first propose an Adversarial Soft kNN (ASK) loss for developing more effective kNN-based deep neural network attack strategies and designing better defense methods against them. Our ASK loss provides a differentiable surrogate of the expected kNN classification error. It is also interpretable as it preserves the mutual information between the perturbed input and the in-class-reference data. We use the ASK loss to design a novel attack method called the ASK-Attack (ASK-Atk), which shows superior attack efficiency and accuracy degradation relative to previous kNN attacks on hidden layers. We then derive an ASK-Defense (ASK-Def) method that optimizes the worst-case ASK training loss. Experiments on CIFAR-10 (ImageNet) show that (i) ASK-Atk achieves $\geq 13\%$ ( $\geq 13\%$ ) improvement in attack success rate over previous kNN attacks, and (ii) ASK-Def outperforms the conventional adversarial training method by $\geq 6.9\%$ ( $\geq 3.5\%$ ) in terms of robustness improvement. Relevant codes are available at https://github.com/wangren09/ASK.

In this paper, we propose a compression-based anomaly detection method for time series and sequence data using a pattern dictionary. The proposed method is capable of learning complex patterns in a training data sequence, using these learned patterns to detect potentially anomalous patterns in a test data sequence. The proposed pattern dictionary method uses a measure of complexity of the test sequence as an anomaly score that can be used to perform stand-alone anomaly detection. We also show that when combined with a universal source coder, the proposed pattern dictionary yields a powerful atypicality detector that is equally applicable to anomaly detection. The pattern dictionary-based atypicality detector uses an anomaly score defined as the difference between the complexity of the test sequence data encoded by the trained pattern dictionary (typical) encoder and the universal (atypical) encoder, respectively. We consider two complexity measures: the number of parsed phrases in the sequence, and the length of the encoded sequence (codelength). Specializing to a particular type of universal encoder, the Tree-Structured Lempel–Ziv (LZ78), we obtain a novel non-asymptotic upper bound, in terms of the Lambert W function, on the number of distinct phrases resulting from the LZ78 parser. This non-asymptotic bound determines the range of anomaly score. As a concrete application, we illustrate the pattern dictionary framework for constructing a baseline of health against which anomalous deviations can be detected.

Pedestrians in videos have a wide range of appearance factors, such as body poses, occlusions, and complex backgrounds, which make their detection difficult. Moreover, a proposal shift problem causes the loss of body parts, such as head and legs in pedestrian detection, which further degrades the detection accuracy. In this paper, we propose part-level convolutional neural networks (CNNs) for pedestrian detection using saliency and boundary box (BB) alignment. The proposed network consists of two subnetworks: detection and alignment. In the detection subnetwork, we use saliency to remove false positives, such as lamp posts and trees, by combining a fully convolutional network and a class activation map to extract deep features. Subsequently, we adopt the BB alignment on detection proposals in the alignment subnetwork to overcome the proposal shift problem by applying the part-level CNN to recall the lost body parts. The experimental results on various datasets demonstrate that the proposed method remarkably improves accuracy in pedestrian detection and outperforms the existing state-of-the-art techniques.

Infection of respiratory mucosa with viral pathogens triggers complex immunologic events in the affected host. We sought to characterize this response through proteomic analysis of nasopharyngeal lavage in human subjects experimentally challenged with influenza A/H3N2 or human rhinovirus, and to develop targeted assays measuring peptides involved in this host response allowing classification of acute respiratory virus infection. Unbiased proteomic discovery analysis identified 3285 peptides corresponding to 438 unique proteins, and revealed that infection with H3N2 induces significant alterations in protein expression. These include proteins involved in acute inflammatory response, innate immune response, and the complement cascade. These data provide insights into the nature of the biological response to viral infection of the upper respiratory tract, and the proteins that are dysregulated by viral infection form the basis of signature that accurately classifies the infected state. Verification of this signature using targeted mass spectrometry in independent cohorts of subjects challenged with influenza or rhinovirus demonstrates that it performs with high accuracy (0.8623 AUROC, 75% TPR, 97.46% TNR). With further development as a clinical diagnostic, this signature may have utility in rapid screening for emerging infections, avoidance of inappropriate antibacterial therapy, and more rapid implementation of appropriate therapeutic and public health strategies.

Scaling up robot swarms to collectives of hundreds or even thousands without sacrificing sensing, processing, and locomotion capabilities is a challenging problem. Low-cost robots are potentially scalable, but the majority of existing systems have limited capabilities, and these limitations substantially constrain the type of experiments that could be performed by robotics researchers. Instead of adding functionality by adding more components and therefore increasing the cost, we demonstrate how low-cost hardware can be used beyond its standard functionality. We systematically review 15 swarm robotic systems and analyse their sensing capabilities by applying a general sensor model from the sensing and measurement community. This work is based on the HoverBot system. A HoverBot is a levitating circuit board that manoeuvres by pulling itself towards magnetic anchors that are embedded into the robot arena. We show that HoverBot’s magnetic field readouts from its Hall-effect sensor can be associated to successful movement, robot rotation and collision measurands. We build a time series classifier based on these magnetic field readouts. We modify and apply signal processing techniques to enable the online classification of the time-variant magnetic field measurements on HoverBot’s low-cost microcontroller. We enabled HoverBot with successful movement, rotation, and collision sensing capabilities by utilising its single Hall-effect sensor. We discuss how our classification method could be applied to other sensors to increase a robot’s functionality while retaining its cost.

Scaling up robot swarms to collectives of hundreds or even thousands without sacrificing sensing, processing, and locomotion capabilities is a challenging problem. Low-cost robots are potentially scalable, but the majority of existing systems have limited capabilities, and these limitations substantially constrain the type of experiments that could be performed by robotics researchers. As an alternative to increasing the quantity of robots by reducing their functionality, we have developed a new technology that delivers increased functionality at low-cost. In this study, we present a comprehensive literature review on the most commonly used locomotion strategies of swarm robotic systems. We introduce a new type of low-friction locomotion—active low-friction locomotion—and we show its first implementation in the HoverBot system. The HoverBot system consists of an air levitation and magnet table, and a HoverBot agent. HoverBot agents are levitating circuit boards that we have equipped with an array of planar coils and a Hall-effect sensor. The HoverBot agent uses its coils to pull itself toward magnetic anchors that are embedded into a levitation table. These robots use active low-friction locomotion; consist of only surface-mount components; circumvent actuator calibration; are capable of odometry by using a single Hall-effect sensor; and perform precise movement. We conducted three hours of experimental evaluation of the HoverBot system in which we observed the system performing more than 10,000 steps. We also demonstrate formation movement, random collision, and straight collisions with two robots. This study demonstrates that active low-friction locomotion is an alternative to wheeled and slip-stick locomotion in the field of swarm robotics.

This paper proposes a geometric estimator of dependency between a pair of multivariate random variables. The proposed estimator of dependency is based on a randomly permuted geometric graph (the minimal spanning tree) over the two multivariate samples. This estimator converges to a quantity that we call the geometric mutual information (GMI), which is equivalent to the Henze−Penrose divergence. between the joint distribution of the multivariate samples and the product of the marginals. The GMI has many of the same properties as standard MI but can be estimated from empirical data without density estimation; making it scalable to large datasets. The proposed empirical estimator of GMI is simple to implement, involving the construction of an minimal spanning tree (MST) spanning over both the original data and a randomly permuted version of this data. We establish asymptotic convergence of the estimator and convergence rates of the bias and variance for smooth multivariate density functions belonging to a Hölder class. We demonstrate the advantages of our proposed geometric dependency estimator in a series of experiments.

Recent work has focused on the problem of nonparametric estimation of information divergence functionals between two continuous random variables. Many existing approaches require either restrictive assumptions about the density support set or difficult calculations at the support set boundary which must be known a priori. The mean squared error (MSE) convergence rate of a leave-one-out kernel density plug-in divergence functional estimator for general bounded density support sets is derived where knowledge of the support boundary, and therefore, the boundary correction is not required. The theory of optimally weighted ensemble estimation is generalized to derive a divergence estimator that achieves the parametric rate when the densities are sufficiently smooth. Guidelines for the tuning parameter selection and the asymptotic distribution of this estimator are provided. Based on the theory, an empirical estimator of Rényi-α divergence is proposed that greatly outperforms the standard kernel density plug-in estimator in terms of mean squared error, especially in high dimensions. The estimator is shown to be robust to the choice of tuning parameters. We show extensive simulation results that verify the theoretical results of our paper. Finally, we apply the proposed estimator to estimate the bounds on the Bayes error rate of a cell classification problem.

This paper introduces a statistical methodology for the identification of differentially expressed genes in DNA microarray experiments based on multiple criteria. These criteria are false discovery rate (FDR), variance-normalized differential expression levels (paired t statistics), and minimum acceptable difference (MAD). The methodology also provides a set of simultaneous FDR confidence intervals on the true expression differences. The analysis can be implemented as a two-stage algorithm in which there is an initial screen that controls only FDR, which is then followed by a second screen which controls both FDR and MAD. It can also be implemented by computing and thresholding the set of FDR P values for each gene that satisfies the MAD criterion. We illustrate the procedure to identify differentially expressed genes from a wild type versus knockout comparison of microarray data.

Predicting the dynamics and functions of microbiomes constructed from the bottom-up is a key challenge in exploiting them to our benefit. Current models based on ecological theory fail to capture complex community behaviors due to higher order interactions, do not scale well with increasing complexity and in considering multiple functions. We develop and apply a long short-term memory (LSTM) framework to advance our understanding of community assembly and health-relevant metabolite production using a synthetic human gut community. A mainstay of recurrent neural networks, the LSTM learns a high dimensional data-driven non-linear dynamical system model. We show that the LSTM model can outperform the widely used generalized Lotka-Volterra model based on ecological theory. We build methods to decipher microbe-microbe and microbe-metabolite interactions from an otherwise black-box model. These methods highlight that Actinobacteria, Firmicutes and Proteobacteria are significant drivers of metabolite production whereas Bacteroides shape community dynamics. We use the LSTM model to navigate a large multidimensional functional landscape to design communities with unique health-relevant metabolite profiles and temporal behaviors. In sum, the accuracy of the LSTM model can be exploited for experimental planning and to guide the design of synthetic microbiomes with target dynamic functions.

The ensemble Kalman filter (EnKF) is a data assimilation technique that uses an ensemble of models, updated with data, to track the time evolution of a usually non-linear system. It does so by using an empirical approximation to the well-known Kalman filter. However, its performance can suffer when the ensemble size is smaller than the state space, as is often necessary for computationally burdensome models. This scenario means that the empirical estimate of the state covariance is not full rank and possibly quite noisy. To solve this problem in this high dimensional regime, we propose a computationally fast and easy to implement algorithm called the penalized ensemble Kalman filter (PEnKF). Under certain conditions, it can be theoretically proven that the PEnKF will be accurate (the estimation error will converge to zero) despite having fewer ensemble members than state dimensions. Further, as contrasted to localization methods, the proposed approach learns the covariance structure associated with the dynamical system. These theoretical results are supported with simulations of several non-linear and high dimensional systems.

There is great potential for host-based gene expression analysis to impact the early diagnosis of infectious diseases. In particular, the influenza pandemic of 2009 highlighted the challenges and limitations of traditional pathogen-based testing for suspected upper respiratory viral infection. We inoculated human volunteers with either influenza A (A/Brisbane/59/2007 (H1N1) or A/Wisconsin/67/2005 (H3N2)), and assayed the peripheral blood transcriptome every 8 hours for 7 days. Of 41 inoculated volunteers, 18 (44%) developed symptomatic infection. Using unbiased sparse latent factor regression analysis, we generated a gene signature (or factor) for symptomatic influenza capable of detecting 94% of infected cases. This gene signature is detectable as early as 29 hours post-exposure and achieves maximal accuracy on average 43 hours (p = 0.003, H1N1) and 38 hours (p-value = 0.005, H3N2) before peak clinical symptoms. In order to test the relevance of these findings in naturally acquired disease, a composite influenza A signature built from these challenge studies was applied to Emergency Department patients where it discriminates between swine-origin influenza A/H1N1 (2009) infected and non-infected individuals with 92% accuracy. The host genomic response to Influenza infection is robust and may provide the means for detection before typical clinical symptoms are apparent.

Exposure to influenza viruses is necessary, but not sufficient, for healthy human hosts to develop symptomatic illness. The host response is an important determinant of disease progression. In order to delineate host molecular responses that differentiate symptomatic and asymptomatic Influenza A infection, we inoculated 17 healthy adults with live influenza (H3N2/Wisconsin) and examined changes in host peripheral blood gene expression at 16 timepoints over 132 hours. Here we present distinct transcriptional dynamics of host responses unique to asymptomatic and symptomatic infections. We show that symptomatic hosts invoke, simultaneously, multiple pattern recognition receptors-mediated antiviral and inflammatory responses that may relate to virus-induced oxidative stress. In contrast, asymptomatic subjects tightly regulate these responses and exhibit elevated expression of genes that function in antioxidant responses and cell-mediated responses. We reveal an ab initio molecular signature that strongly correlates to symptomatic clinical disease and biomarkers whose expression patterns best discriminate early from late phases of infection. Our results establish a temporal pattern of host molecular responses that differentiates symptomatic from asymptomatic infections and reveals an asymptomatic host-unique non-passive response signature, suggesting novel putative molecular targets for both prognostic assessment and ameliorative therapeutic intervention in seasonal and pandemic influenza.