RESUMO
Antimicrobial peptides (AMPs) are promising candidates for new antibiotics due to their broad-spectrum activity against pathogens and reduced susceptibility to resistance development. Deep-learning techniques, such as deep generative models, offer a promising avenue to expedite the discovery and optimization of AMPs. A remarkable example is the Feedback Generative Adversarial Network (FBGAN), a deep generative model that incorporates a classifier during its training phase. Our study aims to explore the impact of enhanced classifiers on the generative capabilities of FBGAN. To this end, we introduce two alternative classifiers for the FBGAN framework, both surpassing the accuracy of the original classifier. The first classifier utilizes the k-mers technique, while the second applies transfer learning from the large protein language model Evolutionary Scale Modeling 2 (ESM2). Integrating these classifiers into FBGAN not only yields notable performance enhancements compared to the original FBGAN but also enables the proposed generative models to achieve comparable or even superior performance to established methods such as AMPGAN and HydrAMP. This achievement underscores the effectiveness of leveraging advanced classifiers within the FBGAN framework, enhancing its computational robustness for AMP de novo design and making it comparable to existing literature.
Assuntos
Peptídeos Antimicrobianos , Peptídeos Antimicrobianos/química , Peptídeos Antimicrobianos/farmacologia , Desenho de Fármacos/métodos , Redes Neurais de Computação , Aprendizado Profundo , AlgoritmosRESUMO
Deep learning models perform remarkably well on many classification tasks recently. The superior performance of deep neural networks relies on the large number of training data, which at the same time must have an equal class distribution in order to be efficient. However, in most real-world applications, the labeled data may be limited with high imbalance ratios among the classes, and thus, the learning process of most classification algorithms is adversely affected resulting in unstable predictions and low performance. Three main categories of approaches address the problem of imbalanced learning, i.e., data-level, algorithmic level, and hybrid methods, which combine the two aforementioned approaches. Data generative methods are typically based on generative adversarial networks, which require significant amounts of data, while model-level methods entail extensive domain expert knowledge to craft the learning objectives, thereby being less accessible for users without such knowledge. Moreover, the vast majority of these approaches are designed and applied to imaging applications, less to time series, and extremely rare to both of them. To address the above issues, we introduce GENDA, a generative neighborhood-based deep autoencoder, which is simple yet effective in its design and can be successfully applied to both image and time-series data. GENDA is based on learning latent representations that rely on the neighboring embedding space of the samples. Extensive experiments, conducted on a variety of widely-used real datasets demonstrate the efficacy of the proposed method. Impact Statement: Imbalanced data classification is an actual and important issue in many real-world learning applications hampering most classification tasks. Fraud detection, biomedical imaging categorizing healthy people versus patients, and object detection are some indicative domains with an economic, social and technological impact, which are greatly affected by inherent imbalanced data distribution. However, the majority of the existing algorithms that address the imbalanced classification problem are designed with a particular application in mind, and thus they can be used with specific datasets and even hyperparameters. The generative model introduced in this paper overcomes this limitation and produces improved results for a large class of imaging and time series data even under severe imbalance ratios, making it quite competitive.
RESUMO
Forecasting the values of essential climate variables like land surface temperature and soil moisture can play a paramount role in understanding and predicting the impact of climate change. This work concerns the development of a deep learning model for analyzing and predicting spatial time series, considering both satellite derived and model-based data assimilation processes. To that end, we propose the Embedded Temporal Convolutional Network (E-TCN) architecture, which integrates three different networks, namely an encoder network, a temporal convolutional network, and a decoder network. The model accepts as input satellite or assimilation model derived values, such as land surface temperature and soil moisture, with monthly periodicity, going back more than fifteen years. We use our model and compare its results with the state-of-the-art model for spatiotemporal data, the ConvLSTM model. To quantify performance, we explore different cases of spatial resolution, spatial region extension, number of training examples and prediction windows, among others. The proposed approach achieves better performance in terms of prediction accuracy, while using a smaller number of parameters compared to the ConvLSTM model. Although we focus on two specific environmental variables, the method can be readily applied to other variables of interest.
Assuntos
Solo , TemperaturaRESUMO
MOTIVATION: Protein structural class prediction is one of the most significant problems in bioinformatics, as it has a prominent role in understanding the function and evolution of proteins. Designing a computationally efficient but at the same time accurate prediction method remains a pressing issue, especially for sequences that we cannot obtain a sufficient amount of homologous information from existing protein sequence databases. Several studies demonstrate the potential of utilizing chaos game representation along with time series analysis tools such as recurrence quantification analysis, complex networks, horizontal visibility graphs (HVG) and others. However, the majority of existing works involve a large amount of features and they require an exhaustive, time consuming search of the optimal parameters. To address the aforementioned problems, this work adopts the generalized multidimensional recurrence quantification analysis (GmdRQA) as an efficient tool that enables to process concurrently a multidimensional time series and reduce the number of features. In addition, two data-driven algorithms, namely average mutual information and false nearest neighbors, are utilized to define in a fast yet precise manner the optimal GmdRQA parameters. RESULTS: The classification accuracy is improved by the combination of GmdRQA with the HVG. Experimental evaluation on a real benchmark dataset demonstrates that our methods achieve similar performance with the state-of-the-art but with a smaller computational cost. AVAILABILITY AND IMPLEMENTATION: The code to reproduce all the results is available at https://github.com/aretiz/protein_structure_classification/tree/main. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
RESUMO
This paper introduces a novel coding/decoding mechanism that mimics one of the most important properties of the human visual system: its ability to enhance the visual perception quality in time. In other words, the brain takes advantage of time to process and clarify the details of the visual scene. This characteristic is yet to be considered by the state-of-the-art quantization mechanisms that process the visual information regardless the duration of time it appears in the visual scene. We propose a compression architecture built of neuroscience models; it first uses the leaky integrate-and-fire (LIF) model to transform the visual stimulus into a spike train and then it combines two different kinds of spike interpretation mechanisms (SIM), the time-SIM and the rate-SIM for the encoding of the spike train. The time-SIM allows a high quality interpretation of the neural code and the rate-SIM allows a simple decoding mechanism by counting the spikes. For that reason, the proposed mechanisms is called Dual-SIM quantizer (Dual-SIMQ). We show that (i) the time-dependency of Dual-SIMQ automatically controls the reconstruction accuracy of the visual stimulus, (ii) the numerical comparison of Dual-SIMQ to the state-of-the-art shows that the performance of the proposed algorithm is similar to the uniform quantization schema while it approximates the optimal behavior of the non-uniform quantization schema and (iii) from the perceptual point of view the reconstruction quality using the Dual-SIMQ is higher than the state-of-the-art.
Assuntos
Compressão de Dados/métodos , Processamento de Imagem Assistida por Computador/métodos , Modelos Neurológicos , Potenciais de Ação/fisiologia , Algoritmos , Humanos , Neurônios/fisiologia , Percepção Visual/fisiologiaRESUMO
Cortical computations are critically reliant on their local circuit, GABAergic cells. In the hippocampus, a large body of work has identified an unprecedented diversity of GABAergic interneurons with pronounced anatomical, molecular, and physiological differences. Yet little is known about the functional properties and activity dynamics of the major hippocampal interneuron classes in behaving animals. Here we use fast, targeted, three-dimensional (3D) two-photon calcium imaging coupled with immunohistochemistry-based molecular identification to retrospectively map in vivo activity onto multiple classes of interneurons in the mouse hippocampal area CA1 during head-fixed exploration and goal-directed learning. We find examples of preferential subtype recruitment with quantitative differences in response properties and feature selectivity during key behavioral tasks and states. These results provide new insights into the collective organization of local inhibitory circuits supporting navigational and mnemonic functions of the hippocampus.
Assuntos
Região CA1 Hipocampal/citologia , Região CA1 Hipocampal/diagnóstico por imagem , Imageamento Tridimensional/métodos , Interneurônios/ultraestrutura , Microscopia de Fluorescência por Excitação Multifotônica/métodos , Animais , Região CA1 Hipocampal/química , Cálcio/análise , Cálcio/metabolismo , Feminino , Interneurônios/química , Masculino , Camundongos , Camundongos Transgênicos , Microscopia Confocal/métodosRESUMO
Monitoring contemporary water distribution networks (WDN) relies increasingly on smart metering technologies and wireless sensor network infrastructures. Smart meters and sensor nodes are deployed to capture and transfer information from the WDN to a control center for further analysis. Due to difficulties in accessing the water assets, many water utility companies employ battery-powered nodes, which restricts the use of high sampling rates, thus limiting the knowledge we can extract from the recorder data. To mitigate this issue, compressive sensing (CS) has been introduced as a powerful framework for reducing dramatically the required bandwidth and storage resources, without diminishing the meaningful information content. Despite its well-established and mathematically rigorous foundations, most of the focus is given on the algorithmic perspective, while the real benefits of CS in practical scenarios are still underexplored. To address this problem, this work investigates the advantages of a CS-based implementation on real sensing devices utilized in smart water networks, in terms of execution speedup and reduced ener experimental evaluation revealed that a CS-based scheme can reduce compression execution times around 50 % , while achieving significant energy savings compared to lossless compression, by selecting a high compression ratio, without compromising reconstruction fidelity. Most importantly, the above significant savings are achieved by simultaneously enabling a weak encryption of the recorded data without the need for additional encryption hardware or software components.
RESUMO
Multispectral sensors constitute a core Earth observation image technology generating massive high-dimensional observations. To address the communication and storage constraints of remote sensing platforms, lossy data compression becomes necessary, but it unavoidably introduces unwanted artifacts. In this work, we consider the encoding of multispectral observations into high-order tensor structures which can naturally capture multi-dimensional dependencies and correlations, and we propose a resource-efficient compression scheme based on quantized low-rank tensor completion. The proposed method is also applicable to the case of missing observations due to environmental conditions, such as cloud cover. To quantify the performance of compression, we consider both typical image quality metrics as well as the impact on state-of-the-art deep learning-based land-cover classification schemes. Experimental analysis on observations from the ESA Sentinel-2 satellite reveals that even minimal compression can have negative effects on classification performance which can be efficiently addressed by our proposed recovery scheme.
RESUMO
Deep Learning, and Deep Neural Networks in particular, have established themselves as the new norm in signal and data processing, achieving state-of-the-art performance in image, audio, and natural language understanding. In remote sensing, a large body of research has been devoted to the application of deep learning for typical supervised learning tasks such as classification. Less yet equally important effort has also been allocated to addressing the challenges associated with the enhancement of low-quality observations from remote sensing platforms. Addressing such channels is of paramount importance, both in itself, since high-altitude imaging, environmental conditions, and imaging systems trade-offs lead to low-quality observation, as well as to facilitate subsequent analysis, such as classification and detection. In this paper, we provide a comprehensive review of deep-learning methods for the enhancement of remote sensing observations, focusing on critical tasks including single and multi-band super-resolution, denoising, restoration, pan-sharpening, and fusion, among others. In addition to the detailed analysis and comparison of recently presented approaches, different research avenues which could be explored in the future are also discussed.
RESUMO
Sensor-based activity recognition is encountered in innumerable applications of the arena of pervasive healthcare and plays a crucial role in biomedical research. Nonetheless, the frequent situation of unobserved measurements impairs the ability of machine learning algorithms to efficiently extract context from raw streams of data. In this paper, we study the problem of accurate estimation of missing multimodal inertial data and we propose a classification framework that considers the reconstruction of subsampled data during the test phase. We introduce the concept of forming the available data streams into low-rank two-dimensional (2-D) and 3-D Hankel structures, and we exploit data redundancies using sophisticated imputation techniques, namely matrix and tensor completion. Moreover, we examine the impact of reconstruction on the classification performance by experimenting with several state-of-the-art classifiers. The system is evaluated with respect to different data structuring scenarios, the volume of data available for reconstruction, and various levels of missing values per device. Finally, the tradeoff between subsampling accuracy and energy conservation in wearable platforms is examined. Our analysis relies on two public datasets containing inertial data, which extend to numerous activities, multiple sensing parameters, and body locations. The results highlight that robust classification accuracy can be achieved through recovery, even for extremely subsampled data streams.
Assuntos
Algoritmos , Atividades Humanas/classificação , Reconhecimento Automatizado de Padrão/métodos , Aprendizado de Máquina Supervisionado , Humanos , Tecnologia de Sensoriamento RemotoRESUMO
Modeling the activity of an ensemble of neurons can provide critical insights into the workings of the brain. In this work we examine if learning based signal modeling can contribute to a high quality modeling of neuronal signal data. To that end, we employ the sparse coding and dictionary learning schemes for capturing the behavior of neuronal responses into a small number of representative prototypical signals. Performance is measured by the reconstruction quality of clean and noisy test signals, which serves as an indicator of the generalization and discrimination capabilities of the learned dictionaries. To validate the merits of the proposed approach, a novel dataset of the actual recordings from 183 neurons from the primary visual cortex of a mouse in early postnatal development was developed and investigated. The results demonstrate that high quality modeling of testing data can be achieved from a small number of training examples and that the learned dictionaries exhibit significant specificity when introducing noise.
RESUMO
This paper introduces a novel framework for compressive sensing of biomedical ultrasonic signals based on modelling data with stable distributions. We propose an approach to â(p) norm minimisation that employs the iteratively reweighted least squares (IRLS) algorithm but in which the parameter p is judiciously chosen by relating it to the characteristic exponent of the underlying alpha-stable distributed data. Our results show that the proposed algorithm, which we prefer to call S ± S-IRLS, outperforms previously proposed â(1) minimisation algorithms, such as basis pursuit or orthogonal matching pursuit, both visually and in terms of PSNR.
Assuntos
Ondas de Rádio , Ultrassonografia , AlgoritmosRESUMO
The aim of this paper is to introduce a novel non-Gaussian statistical model-based approach for medical image fusion based on the Meridian distribution. The paper also includes a new approach to estimate the parameters of generalized Cauchy distribution. The input images are first decomposed using the Dual-Tree Complex Wavelet Transform (DT-CWT) with the subband coefficients modelled as Meridian random variables. Then, the convolution of Meridian distributions is applied as a probabilistic prior to model the fused coefficients, and the weights used to combine the source images are optimised via Maximum Likelihood (ML) estimation. The superior performance of the proposed method is demonstrated using medical images.
Assuntos
Algoritmos , Interpretação de Imagem Assistida por Computador/métodos , Reconhecimento Automatizado de Padrão/métodos , Técnica de Subtração , Interpretação Estatística de Dados , Humanos , Aumento da Imagem/métodos , Funções Verossimilhança , Reprodutibilidade dos Testes , Sensibilidade e Especificidade , Distribuições EstatísticasRESUMO
BACKGROUND: MicroRNAs (miRNAs) are small, single stranded RNAs with a key role in post-transcriptional regulation of thousands of genes across numerous species. While several computational methods are currently available for identifying miRNA genes, accurate prediction of the mature miRNA remains a challenge. Existing approaches fall short in predicting the location of mature miRNAs but also in finding the functional strand(s) of miRNA precursors. METHODOLOGY/PRINCIPAL FINDINGS: Here, we present a computational tool that incorporates a Naive Bayes classifier to identify mature miRNA candidates based on sequence and secondary structure information of their miRNA precursors. We take into account both positive (true mature miRNAs) and negative (same-size non-mature miRNA sequences) examples to optimize sensitivity as well as specificity. Our method can accurately predict the start position of experimentally verified mature miRNAs for both human and mouse, achieving a significantly larger (often double) performance accuracy compared with two existing methods. Moreover, the method exhibits a very high generalization performance on miRNAs from two other organisms. More importantly, our method provides direct evidence about the features of miRNA precursors which may determine the location of the mature miRNA. We find that the triplet of positions 7, 8 and 9 from the mature miRNA end towards the closest hairpin have the largest discriminatory power, are relatively conserved in terms of sequence composition (mostly contain a Uracil) and are located within or in very close proximity to the hairpin loop, suggesting the existence of a possible recognition site for Dicer and associated proteins. CONCLUSIONS: This work describes a novel algorithm for identifying the start position of mature miRNA(s) produced by miRNA precursors. Our tool has significantly better (often double) performance than two existing approaches and provides new insights about the potential use of specific sequence/structural information as recognition signals for Dicer processing. Web Tool available at: http://mirna.imbb.forth.gr/MatureBayes.html.
Assuntos
Algoritmos , Biologia Computacional/métodos , MicroRNAs/metabolismo , Precursores de RNA/metabolismo , Animais , Sequência de Bases , Drosophila melanogaster/genética , Humanos , Camundongos , MicroRNAs/química , MicroRNAs/genética , Probabilidade , Precursores de RNA/química , Precursores de RNA/genética , Curva ROC , Ribonuclease III/metabolismo , Peixe-Zebra/genéticaRESUMO
This paper addresses the construction of a novel efficient rotation-invariant texture retrieval method that is based on the alignment in angle of signatures obtained via a steerable sub-Gaussian model. In our proposed scheme, we first construct a steerable multivariate sub-Gaussian model, where the fractional lower-order moments of a given image are associated with those of its rotated versions. The feature extraction step consists of estimating the so-called covariations between the orientation subbands of the corresponding steerable pyramid at the same or at adjacent decomposition levels and building an appropriate signature that can be rotated directly without the need of rotating the image and recalculating the signature. The similarity measurement between two images is performed using a matrix-based norm that includes a signature alignment in angle between the images being compared, achieving in this way the desired rotation-invariance property. Our experimental results show how this retrieval scheme achieves a lower average retrieval error, as compared to previously proposed methods having a similar computational complexity, while at the same time being competitive with the best currently known state-of-the-art retrieval system. In conclusion, our retrieval method provides the best compromise between complexity and average retrieval performance.
Assuntos
Algoritmos , Inteligência Artificial , Interpretação de Imagem Assistida por Computador/métodos , Reconhecimento Automatizado de Padrão/métodos , Técnica de Subtração , Simulação por Computador , Aumento da Imagem/métodos , Modelos Estatísticos , Distribuição Normal , RotaçãoRESUMO
This paper addresses the task of recovering the geoacoustic parameters of a shallow-water environment using measurements of the acoustic field due to a known source and a neural network based inversion process. First, a novel efficient "observable" of the acoustic signal is proposed, which represents the signal in accordance with the recoverable parameters. Motivated by recent studies in non-Gaussian statistical theory, the observable is defined as a set of estimated model parameters of the alpha-stable distributions, which fit the marginal statistics of the wavelet subband coefficients, obtained after the transformation of the original signal via a one-dimensional wavelet decomposition. Following the modeling process to extract the observables as features, a radial basis functions neural network is employed to approximate the vector function that takes as input the observables and gives as output the corresponding set of environmental parameters. The performance of the proposed approach in recovering the sound speed and density in the substrate of a typical shallow-water environment is evaluated using a database of synthetic acoustic signals, generated by means of a normal-mode acoustic propagation algorithm.
RESUMO
This paper presents a novel rotation-invariant image retrieval scheme based on a transformation of the texture information via a steerable pyramid. First, we fit the distribution of the subband coefficients using a joint alpha-stable sub-Gaussian model to capture their non-Gaussian behavior. Then, we apply a normalization process in order to Gaussianize the coefficients. As a result, the feature extraction step consists of estimating the covariances between the normalized pyramid coefficients. The similarity between two distinct texture images is measured by minimizing a rotation-invariant version of the Kullback-Leibler Divergence between their corresponding multivariate Gaussian distributions, where the minimization is performed over a set of rotation angles.