GNNMF: a multi-view graph neural network for ATAC-seq motif finding.

BACKGROUND: The Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) utilizes the Transposase Tn5 to probe open chromatic, which simultaneously reveals multiple transcription factor binding sites (TFBSs) compared to traditional technologies. Deep learning (DL) technology, including convolutional neural networks (CNNs), has successfully found motifs from ATAC-seq data. Due to the limitation of the width of convolutional kernels, the existing models only find motifs with fixed lengths. A Graph neural network (GNN) can work on non-Euclidean data, which has the potential to find ATAC-seq motifs with different lengths. However, the existing GNN models ignored the relationships among ATAC-seq sequences, and their parameter settings should be improved. RESULTS: In this study, we proposed a novel GNN model named GNNMF to find ATAC-seq motifs via GNN and background coexisting probability. Our experiment has been conducted on 200 human datasets and 80 mouse datasets, demonstrated that GNNMF has improved the area of eight metrics radar scores of 4.92% and 6.81% respectively, and found more motifs than did the existing models. CONCLUSIONS: In this study, we developed a novel model named GNNMF for finding multiple ATAC-seq motifs. GNNMF built a multi-view heterogeneous graph by using ATAC-seq sequences, and utilized background coexisting probability and the iterloss to find different lengths of ATAC-seq motifs and optimize the parameter sets. Compared to existing models, GNNMF achieved the best performance on TFBS prediction and ATAC-seq motif finding, which demonstrates that our improvement is available for ATAC-seq motif finding.

Deep Open-Curve Snake for Discriminative 3D Neuron Tracking.

Open-Curve Snake (OCS) has been successfully used in three-dimensional tracking of neurites. However, it is limited when dealing with noise-contaminated weak filament signals in real-world applications. In addition, its tracking results are highly sensitive to initial seeds and depend only on image gradient-derived forces. To address these issues and boost the canonical OCS tracker to a new level of learnable deep learning algorithms, we present Deep Open-Curve Snake (DOCS), a novel discriminative 3D neuron tracking framework that simultaneously learns a 3D distance-regression discriminator and a 3D deeply-learned tracker under the energy minimization, which can promote each other. In particular, the open curve tracking process in DOCS is formed as convolutional neural network prediction procedures of new deformation fields, stretching directions, and local radii and iteratively updated by minimizing a tractable energy function containing fitting forces and curve length. By sharing the same deep learning architectures in an end-to-end trainable framework, DOCS is able to fully grasp the information available in the volumetric neuronal data to address segmentation, tracing, and reconstruction of complete neuron structures in the wild. We demonstrated the superiority of DOCS by evaluating it on both the BigNeuron and Diadem datasets where consistently state-of-the-art performances were achieved for comparison against current neuron tracing and tracking approaches. Our method improves the average overlap score and distance score about 1.7% and 17% in the BigNeuron challenge data set, respectively, and the average overlap score about 4.1% in the Diadem dataset.

Hyper-Mol: Molecular Representation Learning via Fingerprint-Based Hypergraph.

With the development of artificial intelligence (AI) in the field of drug design and discovery, learning informative representations of molecules is becoming crucial for those AI-driven tasks. In recent years, the graph neural networks (GNNs) have emerged as a preferred choice of deep learning architecture and have been successfully applied to molecular representation learning (MRL). Up-to-date MRL methods directly apply the message passing mechanism on the atom-level attributes (i.e., atoms and bonds) of molecules. However, they neglect latent yet significant hyperstructured knowledge, such as the information of pharmacophore or functional class. Hence, in this paper, we propose Hyper-Mol, a new MRL framework that applies GNNs to encode hypergraph structures of molecules via fingerprint-based features. Hyper-Mol explores the hyperstructured knowledge and the latent relationships of the fingerprint substructures from a hypergraph perspective. The molecular hypergraph generation algorithm is designed to depict the hyperstructured information with the physical and chemical characteristics of molecules. Thus, the fingerprint-level message passing process can encode both the intra-structured and inter-structured information of fingerprint substructures according to the molecular hypergraphs. We evaluate Hyper-Mol on molecular property prediction tasks, and the experimental results on real-world benchmarks show that Hyper-Mol can learn comprehensive hyperstructured knowledge of molecules and is superior to the state-of-the-art baselines.

Dual Attention Triplet Hashing Network for Image Retrieval.

In recent years, learning-based hashing techniques have proven to be efficient for large-scale image retrieval. However, since most of the hash codes learned by deep hashing methods contain repetitive and correlated information, there are some limitations. In this paper, we propose a Dual Attention Triplet Hashing Network (DATH). DATH is implemented with two-stream ConvNet architecture. Specifically, the first neural network focuses on the spatial semantic relevance, and the second neural network focuses on the channel semantic correlation. These two neural networks are incorporated to create an end-to-end trainable framework. At the same time, in order to make better use of label information, DATH combines triplet likelihood loss and classification loss to optimize the network. Experimental results show that DATH has achieved the state-of-the-art performance on benchmark datasets.

Sparse Coding Driven Deep Decision Tree Ensembles for Nucleus Segmentation in Digital Pathology Images.

Automating generalized nucleus segmentation has proven to be non-trivial and challenging in digital pathology. Most existing techniques in the field rely either on deep neural networks or on shallow learning-based cascading models. The former lacks theoretical understanding and tends to degrade performance when only limited amounts of training data are available while the latter often suffers from limitations for generalization. To address these issues, we propose sparse coding driven deep decision tree ensembles (ScD2TE), an easily trained yet powerful representation learning approach with performance highly competitive to deep neural networks in the generalized nucleus segmentation task. We explore the possibility of stacking several layers based on fast convolutional sparse coding-decision tree ensemble pairwise modules and generate a layer-wise encoder-decoder architecture with intra-decoder and inter-encoder dense connectivity patterns. Under this architecture, all the encoders share the same assumption across the different layers to represent images and interact with their decoders to give fast convergence. Compared with deep neural networks, our proposed ScD2TE does not require back-propagation computation and depends on less hyper-parameters. ScD2TE is able to achieve a fast end-to-end pixel-wise training in a layer-wise manner. We demonstrated the superiority of our segmentation method by evaluating it on the multi-disease state and multi-organ dataset where consistently higher performances were obtained for comparison against other state-of-the-art deep learning techniques and cascading methods with various connectivity patterns.

Large-Scale Circuitry Interactions Upon Earthquake Experiences Revealed by Recurrent Neural Networks.

Brain dynamics has recently received increasing interest due to its significant importance in basic and clinical neurosciences. However, due to inherent difficulties in both data acquisition and data analysis methods, studies on large-scale brain dynamics of mouse with local field potential (LFP) recording are very rare. In this paper, we did a series of works on modeling large-scale mouse brain dynamic activities responding to fearful earthquake. Based on LFP recording data from 13 brain regions that are closely related to fear learning and memory and the effective Bayesian connectivity change point model, we divided the response time series into four stages: "Before," "Earthquake," "Recovery," and "After." We first reported the changes in power and theta-gamma coupling during stage transitions. Then, a recurrent neural network model was designed to model the functional dynamics in these thirteen brain regions and six frequency bands in response to the fear stimulus. Interestingly, our results showed that the functional brain connectivities in theta and gamma bands exhibited distinct response processes: in theta band, there is a separated-united-separated alternation in whole-brain connectivity and a low-high-low change in connectivity strength; however, gamma bands have a united-separated-united transition and a high-low-high alternation in connectivity pattern and strength. In general, our results offer a novel perspective in studying functional brain dynamics under fearful stimulus and reveal its relationship to the brain's structural connectivity substrates.

Contour-Seed Pairs Learning-Based Framework for Simultaneously Detecting and Segmenting Various Overlapping Cells/Nuclei in Microscopy Images.

In this paper, we propose a novel contour-seed pairs learning-based framework for robust and automated cell/nucleus segmentation. Automated granular object segmentation in microscopy images has significant clinical importance for pathology grading of the cell carcinoma and gene expression. The focus of the past literature is dominated by either segmenting a certain type of cells/nuclei or simply splitting the clustered objects without contours inference of them. Our method addresses these issues by formulating the detection and segmentation tasks in terms of a unified regression problem, where a cascade sparse regression chain model is trained and then applied to return object locations and entire boundaries of clustered objects. In particular, we first learn a set of online convolutional features in each layer. Then, in the proposed cascade sparse regression chain, with the input from the learned features, we iteratively update the locations and clustered object boundaries until convergence. In this way, the boundary evidences of each individual object can be easily delineated and be further fed to a complete contour inference procedure optimized by the minimum description length principle. For any probe image, our method enables to analyze free-lying and overlapping cells with complex shapes. Experimental results show that the proposed method is very generic and performs well on contour inferences of various cell/nucleus types. Compared with the current segmentation techniques, our approach achieves state-of-the-art performances on four challenging datasets, i.e., the kidney renal cell carcinoma histopathology dataset, Drosophila Kc167 cellular dataset, differential interference contrast red blood cell dataset, and cervical cytology dataset.

Boundary-to-Marker Evidence-Controlled Segmentation and MDL-Based Contour Inference for Overlapping Nuclei.

This paper presents a novel method for automated morphology delineation and analysis of cell nuclei in histopathology images. Combining the initial segmentation information and concavity measurement, the proposed method first segments clusters of nuclei into individual pieces, avoiding segmentation errors introduced by the scale-constrained Laplacian-of-Gaussian filtering. After that a nuclear boundary-to-marker evidence computing is introduced to delineate individual objects after the refined segmentation process. The obtained evidence set is then modeled by the periodic B-splines with the minimum description length principle, which achieves a practical compromise between the complexity of the nuclear structure and its coverage of the fluorescence signal to avoid the underfitting and overfitting results. The algorithm is computationally efficient and has been tested on the synthetic database as well as 45 real histopathology images. By comparing the proposed method with several state-of-the-art methods, experimental results show the superior recognition performance of our method and indicate the potential applications of analyzing the intrinsic features of nuclei morphology.

Specific mutations in the C-terminus domain of HBV surface antigen significantly correlate with low level of serum HBV-DNA in patients with chronic HBV infection.

BACKGROUND: To define HBsAg-mutations correlated with different serum HBV-DNA levels in HBV chronically-infected drug-naive patients. METHODS: This study included 187 patients stratified into the following ranges of serum HBV-DNA:12-2000 IU/ml, 2000-100,000 IU/ml, and >100,000 IU/ml. HBsAg-mutations were associated with HBV-DNA levels by applying a Bayesian-Partitional-Model and Fisher-exact test. Mutant and wild-type HBV genotype-D genomes were expressed in Huh7 cells and HBsAg-production was determined in cell-supernatants at 3 days-post-transfection. RESULTS: Specific HBsAg-mutations (M197T,-S204N-Y206C/H-F220L) were significantly correlated with serum HBV-DNA <2000 IU/ml (posterior-probability>90%, P < 0.05). The presence of Y206C/H and/or F220L was also associated with lower median (IQR) HBsAg-levels and lower median (IQR) transaminases (for HBsAg:250[115-840] IU/ml for Y206C/H and/or F220L versus 4300[640-11,838] IU/ml for wild-type, P = 0.023; for ALT:28[21-40] IU/ml versus 53[34-90] IU/ml, P < 0.001). These mutations were localized in the HBsAg C-terminus, known to be involved in virion and/or HBsAg secretion. The co-occurrence of Y206C + F220L was found significant by cluster-analysis, (P = 0.02). In addition, in an in-vitro model Y206C + F220L determined a 2.8-3.3 fold-reduction of HBsAg-amount released in supernatants compared to single mutants and wt (Y206C + F220L = 5,679 IU/ml; Y206H = 16,305 IU/ml; F220L = 18,368 IU/ml; Y206C = 18,680 IU/ml; wt = 14,280 IU/ml, P < 0.05). CONCLUSIONS: Specific HBsAg-mutations (compartmentalized in the HBsAg C-terminus) correlated with low-serum HBV-DNA and HBsAg-levels. These findings can be important to understand mechanisms underlying low HBV replicative potential including the inactive-carrier state.

Atomic dynamic functional interaction patterns for characterization of ADHD.

Modeling abnormal temporal dynamics of functional interactions in psychiatric disorders has been of great interest in the neuroimaging field, and thus a variety of methods have been proposed so far. However, the temporal dynamics and disease-related abnormalities of functional interactions within specific data-driven discovered subnetworks have been rarely explored yet. In this work, we propose a novel computational framework composed of an effective Bayesian connectivity change point model for modeling functional brain interactions and their dynamics simultaneously and an effective variant of nonnegative matrix factorization for assessing the functional interaction abnormalities within subnetworks. This framework has been applied on the resting state fmagnetic resonance imaging (fMRI) datasets of 23 children with attention-deficit/hyperactivity disorder (ADHD) and 45 normal control (NC) children, and has revealed two atomic functional interaction patterns (AFIPs) discovered for ADHD and another two AFIPs derived for NC. Together, these four AFIPs could be grouped into two pairs, one common pair representing the common AFIPs in ADHD and NC, and the other abnormal pair representing the abnormal AFIPs in ADHD. Interestingly, by comparing the abnormal AFIP pair, two data-driven abnormal functional subnetworks are derived. Strikingly, by evaluating the approximation based on the four AFIPs, all of the ADHD children were successfully differentiated from NCs without any false positive.

Inferring functional interaction and transition patterns via dynamic Bayesian variable partition models.

Multivariate connectivity and functional dynamics have been of wide interest in the neuroimaging field, and a variety of methods have been developed to study functional interactions and dynamics. In contrast, the temporal dynamic transitions of multivariate functional interactions among brain networks, in particular, in resting state, have been much less explored. This article presents a novel dynamic Bayesian variable partition model (DBVPM) that simultaneously considers and models multivariate functional interactions and their dynamics via a unified Bayesian framework. The basic idea is to detect the temporal boundaries of piecewise quasi-stable functional interaction patterns, which are then modeled by representative signature patterns and whose temporal transitions are characterized by finite-state transition machines. Results on both simulated and experimental datasets demonstrated the effectiveness and accuracy of the DBVPM in dividing temporally transiting functional interaction patterns. The application of DBVPM on a post-traumatic stress disorder (PTSD) dataset revealed substantially different multivariate functional interaction signatures and temporal transitions in the default mode and emotion networks of PTSD patients, in comparison with those in healthy controls. This result demonstrated the utility of DBVPM in elucidating salient features that cannot be revealed by static pair-wise functional connectivity analysis.