Sequencing-based microbial count data analysis is a challenging task due to the presence of numerous non-biological zeros, which can impede downstream analysis. To tackle this issue, we introduce two novel approaches, PhyImpute and UniFracImpute, which leverage similar microbial samples to identify and impute non-biological zeros in microbial count data. Our proposed methods utilize the probability of non-biological zeros and phylogenetic trees to estimate sample-to-sample similarity, thus addressing this challenge. To evaluate the performance of our proposed methods, we conduct experiments using both simulated and real microbial data. The results demonstrate that PhyImpute and UniFracImpute outperform existing methods in recovering the zeros and empowering downstream analyses such as differential abundance analysis, and disease status classification.

MOTIVATION: Spatial transcriptomics is a state-of-art technique that allows researchers to study gene expression patterns in tissues over the spatial domain. As a result of technical limitations, the majority of spatial transcriptomics techniques provide bulk data for each sequencing spot. Consequently, in order to obtain high-resolution spatial transcriptomics data, performing deconvolution becomes essential. Most existing deconvolution methods rely on reference data (e.g., single-cell data), which may not be available in real applications. Current reference-free methods encounter limitations due to their dependence on distribution assumptions, reliance on marker genes, or the absence of leveraging histology and spatial information. Consequently, there is a critical need for the development of highly flexible, robust, and user-friendly reference-free deconvolution methods capable of unifying or leveraging case-specific information in the analysis of spatial transcriptomics data. RESULTS: We propose a novel reference-free method based on regularized non-negative matrix factorization (NMF), named Flexible Analysis of Spatial Transcriptomics (FAST), that can effectively incorporate gene expression data, spatial, and histology information into a unified deconvolution framework. Compared to existing methods, FAST imposes fewer distribution assumptions, utilizes the spatial structure information of tissues, and encourages interpretable factorization results. These features enable greater flexibility and accuracy, making FAST an effective tool for deciphering the complex cell-type composition of tissues and advancing our understanding of various biological processes and diseases. Extensive simulation studies have shown that FAST outperforms other existing reference-free methods. In real data applications, FAST is able to uncover the underlying tissue structures and identify the corresponding marker genes.

Human activity recognition (HAR) based on wearable devices has become an active research direction in the field of ubiquitous computing, and has a wide range of Internet of Things (IoT) applications. Unfortunately, it is challenging to obtain large amounts of labeled sensing data, and manual annotation is time-consuming and labor-intensive, making it impossible for the extensive deployment of HAR systems. Consequently, self-supervised learning has emerged to address this challenge by training on unlabeled data. However, traditional contrastive learning fails to simulate more sample diversity problems caused by environmental heterogeneity and sensor heterogeneity. In this paper, we propose a multi-positive sample quantum contrastive learning (MPSQCL) framework. By increasing the positive samples for contrastive learning and leveraging the advantages of quantum machine learning (QML) techniques, the richer features of input samples are extracted to improve the robustness and generalization of the model. Moreover, we design a new contrastive loss function to adapt to multiple positive sample contrastive learning scenarios. Finally, we validate the effectiveness of the proposed framework on several publicly available HAR datasets.

Abstract Hierarchical reinforcement learning (HRL) has achieved remarkable success and significant progress in complex and long-term decision-making problems. However, HRL training typically entails substantial computational costs and an enormous number of samples. One effective approach to tackle this challenge is hierarchical reinforcement learning from demonstrations (HRLfD), which leverages demonstrations to expedite the training process of HRL. The effectiveness of HRLfD is contingent upon the quality of the demonstrations; hence, suboptimal demonstrations may impede efficient learning. To address this issue, this paper proposes a reachability-based reward shaping (RbRS) method to alleviate the negative interference of suboptimal demonstrations for the HRL agent. The novel HRLfD algorithm based on RbRS is named HRLfD-RbRS, which incorporates the RbRS method to enhance the learning efficiency of HRLfD. Moreover, with the help of this method, the learning agent can explore better policies under the guidance of the suboptimal demonstration. We evaluate the proposed HRLfD-RbRS algorithm on various complex robotic tasks, and the experimental results demonstrate that our method outperforms current state-of-the-art HRLfD algorithms.

Recently, wireless sensing techniques have been widely used for Internet of Things (IoT) applications. Unlike traditional device-based sensing, wireless sensing is contactless, pervasive, low cost, and non-invasive, making it highly suitable for relevant IoT applications. However, most existing methods are highly dependent on high-quality datasets, and the minority class will not achieve a satisfactory performance when suffering from a class imbalance problem. In this article, we propose a time–frequency semantic generative adversarial network framework (i.e., TFSemantic) to address the imbalanced classification problem in human activity recognition using radio frequency (RF) signals. Specifically, the TFSemantic framework can learn semantic features from the minority classes and then generate high-quality signals to restore data balance. It includes a data pre-processing module, a semantic extraction module, a semantic distribution module, and a data augmenter module. In the data pre-processing module, we process four different RF datasets (i.e., WiFi, RFID, UWB, and mmWave). We also develop Fourier semantic feature convolution and attention semantic feature embedding methods for the semantic extraction module. A discrete wavelet transform is utilized for reconstructed RF samples in the semantic distribution module. In data augmenter module, we design an associated loss function to achieve effective adversarial training. Finally, we validate the effectiveness of the proposed TFSemantic framework using different RF datasets, which outperforms several state-of-the-art methods.

To implement ubiquitous wireless sensing, the domain shift problem (e.g., different environments, users, devices) for machine learning based approaches should be addressed. Some existing methods are proven to be effective, such as transfer learning and domain adaptation. Meanwhile, quantum machine learning, a combination of quantum computing and machine learning, has attracted much attention. More importantly, quantum transfer learning (QTL) has been successful for certain applications, e.g., image classification. In this paper, we explore a classical to quantum (C2Q) framework to address the domain shift problem in wireless sensing by exploiting the great potential of QTL. Specifically, we first analyze the data shift problem in various types of wireless datasets by calculating the Kullback-Leibler (KL) divergence of different domains. Then, a QTL framework is designed to introduce importance weighting and adversarial strategies. We finally evaluate the proposed framework using the representative human activity recognition task on three wireless sensing datasets. Experimental results demonstrate the feasibility of the framework and its great potential for solving the domain shift problem in wireless sensing.

Video Question Answering (VideoQA) aims to comprehend intricate relationships, actions, and events within video content, as well as the inherent links between objects and scenes, to answer text-based questions accurately. Transferring knowledge from the cross-modal pre-trained model CLIP is a natural approach, but its dual-tower structure hinders fine-grained modality interaction, posing challenges for direct application to VideoQA tasks. To address this issue, we introduce a Language-Guided Visual Aggregation (LGVA) network. It employs CLIP as an effective feature extractor to obtain language-aligned visual features with different granularities and avoids resource-intensive video pre-training. The LGVA network progressively aggregates visual information in a bottom-up manner, focusing on both regional and temporal levels, and ultimately facilitating accurate answer prediction. More specifically, it employs local cross-attention to combine pre-extracted question tokens and region embeddings, pinpointing the object of interest in the question. Then, graph attention is utilized to aggregate regions at the frame level and integrate additional captions for enhanced detail. Following this, global cross-attention is used to merge sentence and frame-level embeddings, identifying the video segment relevant to the question. Ultimately, contrastive learning is applied to optimize the similarities between aggregated visual and answer embeddings, unifying upstream and downstream tasks. Our method conserves resources by avoiding large-scale video pre-training and simultaneously demonstrates commendable performance on the NExT-QA, MSVD-QA, MSRVTT-QA, TGIF-QA, and ActivityNet-QA datasets, even outperforming some end-to-end trained models. Our code is available at https://github.com/ecoxial2007/LGVA_VideoQA.

Neurons, represented as a tree structure of morphology, have various distinguished branches of dendrites. Different types of synaptic receptors distributed over dendrites are responsible for receiving inputs from other neurons. NMDA receptors (NMDARs) are expressed as excitatory units, and play a key physiological role in synaptic function. Although NMDARs are widely expressed in most types of neurons, they play a different role in the cerebellar Purkinje cells (PCs). Utilizing a computational PC model with detailed dendritic morphology, we explored the role of NMDARs at different parts of dendritic branches and regions. We found somatic responses can switch from silent, to simple spikes and complex spikes, depending on specific dendritic branches. Detailed examination of the dendrites regarding their diameters and distance to soma revealed diverse response patterns, yet explain two firing modes, simple and complex spike. Taken together, these results suggest that NMDARs play an important role in controlling excitability sensitivity while taking into account the factor of dendritic properties. Given the complexity of neural morphology varying in cell types, our work suggests that the functional role of NMDARs is not stereotyped but highly interwoven with local properties of neuronal structure.

Synchronous oscillations in neural populations are considered being controlled by inhibitory neurons. In the granular layer of the cerebellum, two major types of cells are excitatory granular cells (GCs) and inhibitory Golgi cells (GoCs). GC spatiotemporal dynamics, as the output of the granular layer, is highly regulated by GoCs. However, there are various types of inhibition implemented by GoCs. With inputs from mossy fibers, GCs and GoCs are reciprocally connected to exhibit different network motifs of synaptic connections. From the view of GCs, feedforward inhibition is expressed as the direct input from GoCs excited by mossy fibers, whereas feedback inhibition is from GoCs via GCs themselves. In addition, there are abundant gap junctions between GoCs showing another form of inhibition. It remains unclear how these diverse copies of inhibition regulate neural population oscillation changes. Leveraging a computational model of the granular layer network, we addressed this question to examine the emergence and modulation of network oscillation using different types of inhibition. We show that at the network level, feedback inhibition is crucial to generate neural oscillation. When short-term plasticity was equipped on GoC-GC synapses, oscillations were largely diminished. Robust oscillations can only appear with additional gap junctions. Moreover, there was a substantial level of cross-frequency coupling in oscillation dynamics. Such a coupling was adjusted and strengthened by GoCs through feedback inhibition. Taken together, our results suggest that the cooperation of distinct types of GoC inhibition plays an essential role in regulating synchronous oscillations of the GC population. With GCs as the sole output of the granular network, their oscillation dynamics could potentially enhance the computational capability of downstream neurons.

The dynamics of cerebellar neuronal networks is controlled by the underlying building blocks of neurons and synapses between them. For which, the computation of Purkinje cells (PCs), the only output cells of the cerebellar cortex, is implemented through various types of neural pathways interactively routing excitation and inhibition converged to PCs. Such tuning of excitation and inhibition, coming from the gating of specific pathways as well as short-term plasticity (STP) of the synapses, plays a dominant role in controlling the PC dynamics in terms of firing rate and spike timing. PCs receive cascade feedforward inputs from two major neural pathways: the first one is the feedforward excitatory pathway from granule cells (GCs) to PCs; the second one is the feedforward inhibition pathway from GCs, via molecular layer interneurons (MLIs), to PCs. The GC-PC pathway, together with short-term dynamics of excitatory synapses, has been a focus over past decades, whereas recent experimental evidence shows that MLIs also greatly contribute to controlling PC activity. Therefore, it is expected that the diversity of excitation gated by STP of GC-PC synapses, modulated by strong inhibition from MLI-PC synapses, can promote the computation performed by PCs. However, it remains unclear how these two neural pathways are interacted to modulate PC dynamics. Here using a computational model of PC network installed with these two neural pathways, we addressed this question to investigate the change of PC firing dynamics at the level of single cell and network. We show that the nonlinear characteristics of excitatory STP dynamics can significantly modulate PC spiking dynamics mediated by inhibition. The changes in PC firing rate, firing phase, and temporal spike pattern, are strongly modulated by these two factors in different ways. MLIs mainly contribute to variable delays in the postsynaptic action potentials of PCs while modulated by excitation STP. Notably, the diversity of synchronization and pause response in the PC network is governed not only by the balance of excitation and inhibition, but also by the synaptic STP, depending on input burst patterns. Especially, the pause response shown in the PC network can only emerge with the interaction of both pathways. Together with other recent findings, our results show that the interaction of feedforward pathways of excitation and inhibition, incorporated with synaptic short-term dynamics, can dramatically regulate the PC activities that consequently change the network dynamics of the cerebellar circuit.

MOTIVATION: Single-cell RNA-sequencing (scRNA-seq) has become an important tool to unravel cellular heterogeneity, discover new cell (sub)types, and understand cell development at single-cell resolution. However, one major challenge to scRNA-seq research is the presence of 'drop-out' events, which usually is due to extremely low mRNA input or the stochastic nature of gene expression. In this article, we present a novel single-cell RNA-seq drop-out correction (scDoc) method, imputing drop-out events by borrowing information for the same gene from highly similar cells. RESULTS: scDoc is the first method that directly involves drop-out information to accounting for cell-to-cell similarity estimation, which is crucial in scRNA-seq drop-out imputation but has not been appropriately examined. We evaluated the performance of scDoc using both simulated data and real scRNA-seq studies. Results show that scDoc outperforms the existing imputation methods in reference to data visualization, cell subpopulation identification and differential expression detection in scRNA-seq data. AVAILABILITY AND IMPLEMENTATION: R code is available at https://github.com/anlingUA/scDoc. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Motivation: The advent of high-throughput next generation sequencing technology has greatly promoted the field of metagenomics where previously unattainable information about microbial communities can be discovered. Detecting differentially abundant features (e.g. species or genes) plays a critical role in revealing the contributors (i.e. pathogens) to the biological or medical status of microbial samples. However, currently available statistical methods lack power in detecting differentially abundant features contrasting different biological or medical conditions, in particular, for time series metagenomic sequencing data. We have proposed a novel procedure, metaDprof, which is built upon a spline-based method assuming heterogeneous error, to meet the challenges of detecting differentially abundant features from metagenomic samples by comparing different biological/medical conditions across time. It contains two stages: (i) global detection on features and (ii) time interval detection for significant features. The detection procedures in both stages are based on sound statistical support. Results: Compared with existing methods the new method metaDprof shows the best performance in comprehensive simulation studies. Not only can it accurately detect features relating to the biological condition or disease status of samples but it also can accurately detect the starting and ending time points when the differences arise. The proposed method is also applied to a real metagenomic dataset and the results provide an interesting angle to understand the relationship between the microbiota in mouse gut and diet type. Availability and Implementation: R code and an example dataset are available at https://cals.arizona.edu/∼anling/sbg/software.htm. Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.

Recent studies on category relations have shown the promising progress in addressing classification problems. Existing works independently consider the known relation and classifier optimization, and thus restrain the room for performance improvement. In this work, a new loss function is proposed to leverage the underlining relations among categories and classifiers. In addition, the bipolar relation (BR) graph is employed to formulate a general form for diverse relations. This bipolar graph is automatically learnt for reliving the constraints which may happen during the cost minimization. Extensive experiments on three benchmarks with various hypotheses and graphs demonstrate that our method can offer a significant performance improvement by jointly learning from both BR graph and hypothesis, in particular on a small training dataset scenario that suffers from severe overfitting problem.

Recently, robust reversible watermarking (RRW) has gained increasing interests and researchers are seeking more stable image features to design watermark embedding and extraction models for local image regions protection. Previous studies show that it is promising to construct local feature regions (FRs) for RRW to handle this problem. However, selecting non-overlapping FRs and evaluating FRs stability for RRW are still challenging. To target this issue, we first construct an undirected weighted graph based on the FRs distribution pattern and then formulate the non-overlapping FRs selection as a weighted maximal clique problem and develop a maximum gain-cost ratio algorithm for its approximately optimal solution. Furthermore, we design reasonable metrics to evaluate the FRs stability in terms of FRs locations and local image content. Extensive experiments demonstrate the effectiveness and efficiency of our work.

MOTIVATION: The high-throughput sequencing technologies have provided a powerful tool to study the microbial organisms living in various environments. Characterizing microbial interactions can give us insights into how they live and work together as a community. Metagonomic data are usually summarized in a compositional fashion due to varying sampling/sequencing depths from one sample to another. We study the co-occurrence patterns of microbial organisms using their relative abundance information. Analyzing compositional data using conventional correlation methods has been shown prone to bias that leads to artifactual correlations. RESULTS: We propose a novel method, regularized estimation of the basis covariance based on compositional data (REBACCA), to identify significant co-occurrence patterns by finding sparse solutions to a system with a deficient rank. To be specific, we construct the system using log ratios of count or proportion data and solve the system using the l1-norm shrinkage method. Our comprehensive simulation studies show that REBACCA (i) achieves higher accuracy in general than the existing methods when a sparse condition is satisfied; (ii) controls the false positives at a pre-specified level, while other methods fail in various cases and (iii) runs considerably faster than the existing comparable method. REBACCA is also applied to several real metagenomic datasets. AVAILABILITY AND IMPLEMENTATION: The R codes for the proposed method are available at http://faculty.wcas.northwestern.edu/∼hji403/REBACCA.htm CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: With the advance of new sequencing technologies producing massive short reads data, metagenomics is rapidly growing, especially in the fields of environmental biology and medical science. The metagenomic data are not only high dimensional with large number of features and limited number of samples but also complex with a large number of zeros and skewed distribution. Efficient computational and statistical tools are needed to deal with these unique characteristics of metagenomic sequencing data. In metagenomic studies, one main objective is to assess whether and how multiple microbial communities differ under various environmental conditions. RESULTS: We propose a two-stage statistical procedure for selecting informative features and identifying differentially abundant features between two or more groups of microbial communities. In the functional analysis of metagenomes, the features may refer to the pathways, subsystems, functional roles and so on. In the first stage of the proposed procedure, the informative features are selected using elastic net as reducing the dimension of metagenomic data. In the second stage, the differentially abundant features are detected using generalized linear models with a negative binomial distribution. Compared with other available methods, the proposed approach demonstrates better performance for most of the comprehensive simulation studies. The new method is also applied to two real metagenomic datasets related to human health. Our findings are consistent with those in previous reports. AVAILABILITY: R code and two example datasets are available at http://cals.arizona.edu/∼anling/software.htm. SUPPLEMENTARY INFORMATION: Supplementary file is available at Bioinformatics online.

MOTIVATION: The analysis of differential abundance for features (e.g. species or genes) can provide us with a better understanding of microbial communities, thus increasing our comprehension and understanding of the behaviors of microbial communities. However, it could also mislead us about the characteristics of microbial communities if the abundances or counts of features on different scales are not properly normalized within and between communities, prior to the analysis of differential abundance. Normalization methods used in the differential analysis typically try to adjust counts on different scales to a common scale using the total sum, mean or median of representative features across all samples. These methods often yield undesirable results when the difference in total counts of differentially abundant features (DAFs) across different conditions is large. RESULTS: We develop a novel method, Ratio Approach for Identifying Differential Abundance (RAIDA), which utilizes the ratio between features in a modified zero-inflated lognormal model. RAIDA removes possible problems associated with counts on different scales within and between conditions. As a result, its performance is not affected by the amount of difference in total abundances of DAFs across different conditions. Through comprehensive simulation studies, the performance of our method is consistently powerful, and under some situations, RAIDA greatly surpasses other existing methods. We also apply RAIDA on real datasets of type II diabetes and find interesting results consistent with previous reports. AVAILABILITY AND IMPLEMENTATION: An R package for RAIDA can be accessed from http://cals.arizona.edu/%7Eanling/sbg/software.htm.

BACKGROUND: Metagenomics has a great potential to discover previously unattainable information about microbial communities. An important prerequisite for such discoveries is to accurately estimate the composition of microbial communities. Most of prevalent homology-based approaches utilize solely the results of an alignment tool such as BLAST, limiting their estimation accuracy to high ranks of the taxonomy tree. RESULTS: We developed a new homology-based approach called Taxonomic Analysis by Elimination and Correction (TAEC), which utilizes the similarity in the genomic sequence in addition to the result of an alignment tool. The proposed method is comprehensively tested on various simulated benchmark datasets of diverse complexity of microbial structure. Compared with other available methods designed for estimating taxonomic composition at a relatively low taxonomic rank, TAEC demonstrates greater accuracy in quantification of genomes in a given microbial sample. We also applied TAEC on two real metagenomic datasets, oral cavity dataset and Crohn's disease dataset. Our results, while agreeing with previous findings at higher ranks of the taxonomy tree, provide accurate estimation of taxonomic compositions at the species/strain level, narrowing down which species/strains need more attention in the study of oral cavity and the Crohn's disease. CONCLUSIONS: By taking account of the similarity in the genomic sequence TAEC outperforms other available tools in estimating taxonomic composition at a very low rank, especially when closely related species/strains exist in a metagenomic sample.

Content-based video copy detection (CBVCD) has attracted increasing attention in recent years. However, video content description and search efficiency are still two challenges in this domain. To cope with these two problems, this paper proposes a novel CBVCD approach with similarity preserving multimodal hash learning (SPM2H). The pre-processed video keyframes are represented as multiple features from different perspectives. SPM2H integrates the multimodal feature fusion and the hashing function learning into a joint framework. Mapping video keyframes into hash codes can conducts fast similarity search in the Hamming space. The experiments show that our approach achieves good performance in accuracy as well as efficiency.

Robust reversible watermarking (RRW) methods are popular in multimedia for protecting copyright, while preserving intactness of host images and providing robustness against unintentional attacks. However, conventional RRW methods are not readily applicable in practice. That is mainly because 1) they fail to offer satisfactory reversibility on large-scale image datasets; 2) they have limited robustness in extracting watermarks from the watermarked images destroyed by different unintentional attacks; and 3) some of them suffer from extremely poor invisibility for watermarked images. Therefore, it is necessary to have a framework to address these three problems, and further improve its performance. This paper presents a novel pragmatic framework, wavelet-domain statistical quantity histogram shifting and clustering (WSQH-SC). Compared with conventional methods, WSQH-SC ingeniously constructs new watermark embedding and extraction procedures by histogram shifting and clustering, which are important for improving robustness and reducing run-time complexity. Additionally, WSQH-SC includes the property inspired pixel adjustment (PIPA) to effectively handle overflow and underflow of pixels. This results in satisfactory reversibility and invisibility. Furthermore, to increase its practical applicability, WSQH-SC designs an enhanced pixel-wise masking (EPWM) to balance robustness and invisibility. We perform extensive experiments over natural, medical, and synthetic aperture radar (SAR) images to show the effectiveness of WSQH-SC by comparing with the histogram rotation (HR)-based and histogram distribution constrained (HDC) methods.

Histogram-based lossless data embedding (LDE) has been recognized as an effective and efficient way for copyright protection of multimedia. Recently, a LDE method using the statistical quantity histogram has achieved good performance, which utilizes the similarity of the arithmetic average of difference histogram (AADH) to reduce the diversity of images and ensure the stable performance of LDE. However, this method is strongly dependent on some assumptions, which limits its applications in practice. In addition, the capacities of the images with the flat AADH, e.g., texture images, are a little bit low. For this purpose, we develop a novel framework for LDE by incorporating the merits from the generalized statistical quantity histogram (GSQH) and the histogram-based embedding. Algorithmically, we design the GSQH driven LDE framework carefully so that it: (1) utilizes the similarity and sparsity of GSQH to construct an efficient embedding carrier, leading to a general and stable framework; (2) is widely adaptable for different kinds of images, due to the usage of the divide-and-conquer strategy; (3) is scalable for different capacity requirements and avoids the capacity problems caused by the flat histogram distribution; (4) is conditionally robust against JPEG compression under a suitable scale factor; and (5) is secure for copyright protection because of the safe storage and transmission of side information. Thorough experiments over three kinds of images demonstrate the effectiveness of the proposed framework.