Predicting microbe-drug associations with structure-enhanced contrastive learning and self-paced negative sampling strategy.

MOTIVATION: Predicting the associations between human microbes and drugs (MDAs) is one critical step in drug development and precision medicine areas. Since discovering these associations through wet experiments is time-consuming and labor-intensive, computational methods have already been an effective way to tackle this problem. Recently, graph contrastive learning (GCL) approaches have shown great advantages in learning the embeddings of nodes from heterogeneous biological graphs (HBGs). However, most GCL-based approaches don't fully capture the rich structure information in HBGs. Besides, fewer MDA prediction methods could screen out the most informative negative samples for effectively training the classifier. Therefore, it still needs to improve the accuracy of MDA predictions. RESULTS: In this study, we propose a novel approach that employs the Structure-enhanced Contrastive learning and Self-paced negative sampling strategy for Microbe-Drug Association predictions (SCSMDA). Firstly, SCSMDA constructs the similarity networks of microbes and drugs, as well as their different meta-path-induced networks. Then SCSMDA employs the representations of microbes and drugs learned from meta-path-induced networks to enhance their embeddings learned from the similarity networks by the contrastive learning strategy. After that, we adopt the self-paced negative sampling strategy to select the most informative negative samples to train the MLP classifier. Lastly, SCSMDA predicts the potential microbe-drug associations with the trained MLP classifier. The embeddings of microbes and drugs learning from the similarity networks are enhanced with the contrastive learning strategy, which could obtain their discriminative representations. Extensive results on three public datasets indicate that SCSMDA significantly outperforms other baseline methods on the MDA prediction task. Case studies for two common drugs could further demonstrate the effectiveness of SCSMDA in finding novel MDA associations. AVAILABILITY: The source code is publicly available on GitHub https://github.com/Yue-Yuu/SCSMDA-master.

ILGBMSH: an interpretable classification model for the shRNA target prediction with ensemble learning algorithm.

Short hairpin RNA (shRNA)-mediated gene silencing is an important technology to achieve RNA interference, in which the design of potent and reliable shRNA molecules plays a crucial role. However, efficient shRNA target selection through biological technology is expensive and time consuming. Hence, it is crucial to develop a more precise and efficient computational method to design potent and reliable shRNA molecules. In this work, we present an interpretable classification model for the shRNA target prediction using the Light Gradient Boosting Machine algorithm called ILGBMSH. Rather than utilizing only the shRNA sequence feature, we extracted 554 biological and deep learning features, which were not considered in previous shRNA prediction research. We evaluated the performance of our model compared with the state-of-the-art shRNA target prediction models. Besides, we investigated the feature explanation from the model's parameters and interpretable method called Shapley Additive Explanations, which provided us with biological insights from the model. We used independent shRNA experiment data from other resources to prove the predictive ability and robustness of our model. Finally, we used our model to design the miR30-shRNA sequences and conducted a gene knockdown experiment. The experimental result was perfectly in correspondence with our expectation with a Pearson's coefficient correlation of 0.985. In summary, the ILGBMSH model can achieve state-of-the-art shRNA prediction performance and give biological insights from the machine learning model parameters.

Methods for predicting single-cell miRNA in breast cancer.

It has been demonstrated that miRNAs are involved in many biological processes including cell proliferation and differentiation, apoptosis, and stress responses. Although single-cell RNA sequencing technology is prevailing nowadays, it still remains challenging in quantifying miRNA at the single-cell level. Herein, we present the computational methods to infer the single-cell miRNA expression level using its target gene abundances. Firstly, we developed an enrichment-based approach in estimating miRNA expression considering miRNA-mRNA regulation information and miRNA-mRNA correlation signal captured from existing TCGA datasets. Further efforts were made to infer the miRNA expression with machine learning models. The methods were applied to compare the accuracy and robustness with the simulated single-cell data. Finally, we applied the method in single-cell RNA-seq triple negative breast cancer (TNBC) patients to further discover miRNA marker at the single-cell level for the malignant cells. Our tool is available online at: https://github.com/ChengkuiZhao/Single-cell-miRNA-prediction.

Advances and Challenges in De Novo Drug Design Using Three-Dimensional Deep Generative Models.

A persistent goal for de novo drug design is to generate novel chemical compounds with desirable properties in a labor-, time-, and cost-efficient manner. Deep generative models provide alternative routes to this goal. Numerous model architectures and optimization strategies have been explored in recent years, most of which have been developed to generate two-dimensional molecular structures. Some generative models aiming at three-dimensional (3D) molecule generation have also been proposed, gaining attention for their unique advantages and potential to directly design drug-like molecules in a target-conditioning manner. This review highlights current developments in 3D molecular generative models combined with deep learning and discusses future directions for de novo drug design.

Neuropeptide Y1 Receptor Antagonist Alters Gut Microbiota and Alleviates the Ovariectomy-Induced Osteoporosis in Rats.

A plethora of evidence has suggested that gut microbiota is involved in the occurrence and development of postmenopausal osteoporosis (PMO). It has been suggested that neuropeptide Y (NPY) modulates the bone metabolism through Y1 receptor (Y1R), and might be associated with gut microbiota. The present study aims to evaluate the anti-osteoporotic effects of Y1R antagonist and to investigate the potential mechanism by which Y1R antagonist regulates gut microbiota. In this study, eighteen female rats were randomly divided into three groups: the sham surgery (SHAM) group, the ovariectomized (OVX) group, and OVX+BIBO3304 group. After 6 weeks following surgery, Y1R antagonist BIBO3304 was administered to the rats in OVX+BIBO3304 group for 7 days. The bone microstructure and serum biochemical parameters were measured at 12 weeks after operation. The differences in the gut microbiota were analyzed by 16S rDNA gene sequencing. Heat-map and Spearman's correlation analyses were constructed to investigate the correlations between microbiota and bone metabolism-related parameters. The results indicated that OVX+BIBO3304 group showed significantly higher BMD, BV/TV, Tb.Th, Tb.N, Conn.D, and serum Ca2+ level than those in OVX group. Additionally, Y1R antagonist changed the gut microbiota composition with lower Firmicutes/Bacteroidetes ratio and higher proportions of some probiotics, including Lactobacillus. The correlation analysis showed that the changes of gut microbiota were closely associated with bone microstructure and serum Ca2+ levels. Our results suggested that Y1R antagonist played an anti-osteoporotic effect and regulated gut microbiota in OVX rats, indicating the potential to utilize Y1R antagonist as a novel treatment for PMO.

Long Non-Coding RNA CRNDE Regulates Angiogenesis in Hepatoblastoma by Targeting the MiR-203/VEGFA Axis.

OBJECTIVE: MiR-203 has been shown to participate in multiple malignancies, but the role of miR-203 in hepatoblastoma (HB) remains unclear. The aim of our study was to investigate the effects of miR-203 in HB. METHODS: A total of 15 pairs of HB tissues and para-tumour normal tissues were collected for the experiments. RT-qPCR and Western blotting were performed to detect the expression of CRNDE, miR-203, and VEGFA at the mRNA and/or protein levels, respectively. A dual luciferase assay verified the target relationship between miR-203 and the 3'UTR of VEGFA as well as miR-203 and CRNDE. In addition, MTT, wound healing, and tube formation assays were performed to assess the effects of miR-203, VEGFA, and CRNDE on cell proliferation, migration, and angiogenesis, respectively. RESULTS: Our data revealed that miR-203 expression was decreased in HB tissues, while long non-coding RNA (lncRNA) CRNDE expression was increased. The dysregulation of miR-203 and CRNDE was closely related to tumour size and stage. Moreover, overexpression of miR-203 inhibited angiogenesis. A dual luciferase assay verified that VEGFA is a direct target of miR-203 and that CRNDE binds to miR-203. Furthermore, our results showed that miR-203 suppressed cell viability, migration, and angiogenesis by regulating VEGFA expression. Additionally, it was confirmed that CRNDE promoted angiogenesis by negatively regulating miR-203 expression. CONCLUSION: lncRNA CRNDE targets the miR-203/VEGFA axis and promotes angiogenesis in HB. These results provide insight into the underlying mechanisms of HB and indicate that CRNDE and miR-203 might be potential targets for HB therapy.

Extended Target Marginal Distribution Poisson Multi-Bernoulli Mixture Filter.

The existence of clutter, unknown measurement sources, unknown number of targets, and undetected probability are problems for multi-extended target tracking, to address these problems; this paper proposes a gamma-Gaussian-inverse Wishart (GGIW) implementation of a marginal distribution Poisson multi-Bernoulli mixture (MD-PMBM) filter. Unlike existing multiple extended target tracking filters, the GGIW-MD-PMBM filter computes the marginal distribution (MD) and the existence probability of each target, which can shorten the computing time while maintaining good tracking results. The simulation results confirm the validity and reliability of the GGIW-MD-PMBM filter.

A Survey on Secure Computation Based on Homomorphic Encryption in Vehicular Ad Hoc Networks.

In vehicular ad hoc networks (VANETs), the security and privacy of vehicle data are core issues. In order to analyze vehicle data, they need to be computed. Encryption is a common method to guarantee the security of vehicle data in the process of data dissemination and computation. However, encrypted vehicle data cannot be analyzed easily and flexibly. Because homomorphic encryption supports computations of the ciphertext, it can completely solve this problem. In this paper, we provide a comprehensive survey of secure computation based on homomorphic encryption in VANETs. We first describe the related definitions and the current state of homomorphic encryption. Next, we present the framework, communication domains, wireless access technologies and cyber-security issues of VANETs. Then, we describe the state of the art of secure basic operations, data aggregation, data query and other data computation in VANETs. Finally, several challenges and open issues are discussed for future research.

In Vivo Deep-Brain Structural and Hemodynamic Multiphoton Microscopy Enabled by Quantum Dots.

Visualizing deep-brain vasculature and hemodynamics is key to understanding brain physiology and pathology. Among the various adopted imaging modalities, multiphoton microscopy (MPM) is well-known for its deep-brain structural and hemodynamic imaging capability. However, the largest imaging depth in MPM is limited by signal depletion in the deep brain. Here we demonstrate that quantum dots are an enabling material for significantly deeper structural and hemodynamic MPM in mouse brain in vivo. We characterized both three-photon excitation and emission parameters for quantum dots: the measured three-photon cross sections of quantum dots are 4-5 orders of magnitude larger than those of conventional fluorescent dyes excited at the 1700 nm window, while the three-photon emission spectrum measured in the circulating blood in vivo shows a slight red shift and broadening compared with ex vivo measurement. On the basis of these measured results, we further demonstrate both structural and hemodynamic three-photon microscopy in the mouse brain in vivo labeled by quantum dots, at record depths among all MPM modalities at all demonstrated excitation wavelengths.

MicroRNA-876-5p inhibits cell proliferation, migration and invasion by targeting c-Met in osteosarcoma.

Recently, aberrant expression of miR-876-5p has been reported to participate in the progression of several human cancers. However, the expression and function of miR-876-5p in osteosarcoma (OS) are still unknown. Here, we found that the expression of miR-876-5p was significantly down-regulated in OS tissues compared to para-cancerous tissues. Clinical association analysis indicated that underexpression of miR-876-5p was positively correlated with advanced clinical stage and poor differentiation. More importantly, OS patients with low miR-876-5p level had a significant shorter overall survival compared to miR-876-5p high-expressing patients. In addition, gain- and loss-of-function experiments demonstrated that miR-876-5p restoration suppressed whereas miR-876-5p knockdown promoted cell proliferation, migration and invasion in both U2OS and MG63 cells. In vivo studies revealed that miR-876-5p overexpression inhibited tumour growth of OS in mice. Mechanistically, miR-876-5p reduced c-Met abundance in OS cells and inversely correlated c-Met expression in OS tissues. Herein, c-Met was recognized as a direct target of miR-876-5p using luciferase reporter assay. Notably, c-Met restoration rescued miR-876-5p attenuated the proliferation, migration and invasion of OS cells. In conclusion, these findings indicate that miR-876-5p may be used as a potential therapeutic target and promising biomarker for the diagnosis and prognosis of OS.

High-Throughput and Direct Sample Screening Using a Laser Spray Ionization Miniature Mass Spectrometer.

Over the past two decades, great efforts have been made toward mass spectrometer instrument miniaturization. With increasing analytical performances, miniature mass spectrometers are on the edge of being applied to more application scenarios. Besides sensitivity, mass resolution, and instrument portability, high-throughput and little or no sample preparation are also critical features in practical applications. In this study, we report the development of a miniature mass spectrometry (MS) system equipped with a 2D moving platform and a laser spray ionization (LSI) source. The method to make a patterned sample holder was also introduced and optimized for automatic high-throughput sample analyses. With the LSI source, analytes in complex matrix could be directly mass analyzed; in addition to the 2D moving platform, different samples could be analyzed in a high-throughput fashion. Results show that good linearity of quantitation could be achieved for multiple samples. Tens of nanograms of drugs, peptides, and vitamin B could be identified in diluted whole blood samples, and it takes 10 s on average to scan one sample.

A Novel FEM Based T-S Fuzzy Particle Filtering for Bearings-Only Maneuvering Target Tracking.

In this paper, we propose a novel fuzzy expectation maximization (FEM) based Takagi-Sugeno (T-S) fuzzy particle filtering (FEMTS-PF) algorithm for a passive sensor system. In order to incorporate target spatial-temporal information into particle filtering, we introduce a T-S fuzzy modeling algorithm, in which an improved FEM approach is proposed to adaptively identify the premise parameters, and the model probability is adjusted by the premise membership functions. In the proposed FEM, the fuzzy parameter is derived by the fuzzy C-regressive model clustering method based on entropy and spatial-temporal characteristics, which can avoid the subjective influence caused by the artificial setting of the initial value when compared to the traditional FEM. Furthermore, using the proposed T-S fuzzy model, the algorithm samples particles, which can effectively reduce the particle degradation phenomenon and the parallel filtering, can realize the real-time performance of the algorithm. Finally, the results of the proposed algorithm are evaluated and compared to several existing filtering algorithms through a series of Monte Carlo simulations. The simulation results demonstrate that the proposed algorithm is more precise, robust and that it even has a faster convergence rate than the interacting multiple model unscented Kalman filter (IMMUKF), interacting multiple model extended Kalman filter (IMMEKF) and interacting multiple model Rao-Blackwellized particle filter (IMMRBPF).

A Robust Multi-Sensor Data Fusion Clustering Algorithm Based on Density Peaks.

In this paper, a novel multi-sensor clustering algorithm, based on the density peaks clustering (DPC) algorithm, is proposed to address the multi-sensor data fusion (MSDF) problem. The MSDF problem is raised in the multi-sensor target detection (MSTD) context and corresponds to clustering observations of multiple sensors, without prior information on clutter. During the clustering process, the data points from the same sensor cannot be grouped into the same cluster, which is called the cannot link (CL) constraint; the size of each cluster should be within a certain range; and overlapping clusters (if any) must be divided into multiple clusters to satisfy the CL constraint. The simulation results confirm the validity and reliability of the proposed algorithm.

Tracking the Turn Maneuvering Target Using the Multi-Target Bayes Filter with an Adaptive Estimation of Turn Rate.

Tracking the target that maneuvers at a variable turn rate is a challenging problem. The traditional solution for this problem is the use of the switching multiple models technique, which includes several dynamic models with different turn rates for matching the motion mode of the target at each point in time. However, the actual motion mode of a target at any time may be different from all of the dynamic models, because these models are usually limited. To address this problem, we establish a formula for estimating the turn rate of a maneuvering target. By applying the estimation method of the turn rate to the multi-target Bayes (MB) filter, we develop a MB filter with an adaptive estimation of the turn rate, in order to track multiple maneuvering targets. Simulation results indicate that the MB filter with an adaptive estimation of the turn rate, is better than the existing filter at tracking the target that maneuvers at a variable turn rate.

Auxiliary Truncated Unscented Kalman Filtering for Bearings-Only Maneuvering Target Tracking.

Novel auxiliary truncated unscented Kalman filtering (ATUKF) is proposed for bearings-only maneuvering target tracking in this paper. In the proposed algorithm, to deal with arbitrary changes in motion models, a modified prior probability density function (PDF) is derived based on some auxiliary target characteristics and current measurements. Then, the modified prior PDF is approximated as a Gaussian density by using the statistical linear regression (SLR) to estimate the mean and covariance. In order to track bearings-only maneuvering target, the posterior PDF is jointly estimated based on the prior probability density function and the modified prior probability density function, and a practical algorithm is developed. Finally, compared with other nonlinear filtering approaches, the experimental results of the proposed algorithm show a significant improvement for both the univariate nonstationary growth model (UNGM) case and bearings-only target tracking case.

100 W dissipative soliton resonances from a thulium-doped double-clad all-fiber-format MOPA system.

In this paper, we first achieve nanosecond-scale dissipative soliton resonance (DSR) generation in a thulium-doped double-clad fiber (TDF) laser with all-anomalous-dispersion regime, and also first scale the average power up to 100.4 W by employing only two stage TDF amplifiers, corresponding to gains of 19.3 and 14.4 dB, respectively. It is noted that both the fiber laser oscillator and the amplification system employ double-clad fiber as the gain medium for utilizing the advantages in high-gain-availability, high-power-handling and good-mode-quality-maintaining. DSR mode-locking of the TDF oscillator is realized by using a nonlinear optical loop mirror (NOLM), which exhibits all-fiber-format, high nonlinear and passive saturable absorption properties. The TDF oscillator can deliver rectangular-shape pulses with duration ranging from ~3.74 to ~72.19 ns while maintaining a nearly equal output peak power level of ~0.56 W, namely peak power clamping (PPC) effect. Comparatively, the two stage amplifiers can scale the seeding pulses to similar average power levels, but to dramatically different peak powers ranging from ~0.94 to ~18.1 kW depending on the durations. Our TDF master-oscillator-power-amplifier (MOPA) system can provide a high power 2-µm band all-fiber-format laser source both tunable in pulse duration and peak power.

Patellar non-eversion in primary TKA reduces the complication rate.

PURPOSE: This study was designed to evaluate the isolated benefits of patellar non-eversion in total knee arthroplasty (TKA). METHODS: This systematic review and meta-analysis was conducted following the PRISMA statement. A comprehensive search of the MEDLINE/PubMed, Cochrane Library, and Embase databases was performed in August 2014. Randomized controlled trials (RCTs) that considered the handling of the patella as the only variable were included in our review. Quality assessment of RCTs was performed according to the CONSORT statement. The meta-analysis was performed to pool the available data for some parameters. RESULTS: The searches of the MEDLINE/PubMed, Cochrane Library, and Embase databases yielded 10 RCTs, and five RCTs were selected for inclusion in the review. This results suggested that tourniquet time [mean difference (MD) = -5.69; 95% confidence interval (CI) -9.77 to -1.60], length of hospitalization (MD = 1.24; 95% CI 0.54-1.94) and the incidence of complications [odds ratio (OR) = 2.23; 95% CI 1.12-4.44] differed significantly between the eversion group and non-eversion group. No differences in postoperative pain, alignment, and the Insall-Salvati ratio were observed between the groups. CONCLUSION: The patellar non-eversion approach offers a shorter length of hospitalization and lower incidence of postoperative complications, but requires more operative time. The merits of patellar non-eversion for recovery of knee function remain controversial, and more high-quality RCTs are needed to draw clear conclusions. In general, avoidance of patellar eversion is recommended when exposing the knee joint for TKA.

Role of Toll-Like Receptor Signaling in the Pathogenesis of Graft-versus-Host Diseases.

Graft-versus-host disease (GVHD) and infection are major complications after allogeneic hematopoietic stem cell transplantation (allo-HSCT) and the leading causes of morbidity and mortality in HSCT patients. Recent work has demonstrated that the two complications are interdependent. GVHD occurs when allo-reactive donor T lymphocytes are activated by major histocompatibility antigens or minor histocompatibility antigens on host antigen-presenting cells (APCs), with the eventual attack of recipient tissues or organs. Activation of APCs is important for the priming of GVHD and is mediated by innate immune signaling pathways. Current evidence indicates that intestinal microbes and innate pattern-recognition receptors (PRRs) on host APCs, including both Toll-like receptors (TLRs) and nucleotide oligomerization domain (NOD)-like receptors (NLRs), are involved in the pathogenesis of GVHD. Patients undergoing chemotherapy and/or total body irradiation before allo-HSCT are susceptible to aggravated gastrointestinal epithelial cell damage and the subsequent translocation of bacterial components, followed by the release of endogenous dangerous molecules, termed pathogen-associated molecular patterns (PAMPs), which then activate the PRRs on host APCs to trigger local or systemic inflammatory responses that modulate T cell allo-reactivity against host tissues, which is equivalent to GVHD. In other words, infection can, to some extent, accelerate the progression of GVHD. Therefore, the intestinal flora's PAMPs can interact with TLRs to activate and mature APCs, subsequently activate donor T cells with the release of pro-inflammatory cytokines, and eventually, induce GVHD. In the present article, we summarize the current perspectives on the understanding of different TLR signaling pathways and their involvement in the occurrence of GVHD.

GT-CAM: Game Theory based Class Activation Map for GCN.

Graph Convolutional Networks (GCN) have shown outstanding performance in skeleton-based behavior recognition. However, their opacity hampers further development. Researches on the explainability of deep learning have provided solutions to this issue, with Class Activation Map (CAM) algorithms being a class of explainable methods. However, existing CAM algorithms applies to GCN often independently compute the contribution of individual nodes, overlooking the interactions between nodes in the skeleton. Therefore, we propose a game theory based class activation map for GCN (GT-CAM). Firstly, GT-CAM integrates Shapley values with gradient weights to calculate node importance, producing an activation map that highlights the critical role of nodes in decision-making. It also reveals the cooperative dynamics between nodes or local subgraphs for a more comprehensive explanation. Secondly, to reduce the computational burden of Shapley values, we propose a method for calculating Shapley values of node coalitions. Lastly, to evaluate the rationality of coalition partitioning, we propose a rationality evaluation method based on bipartite game interaction and cooperative game theory. Additionally, we introduce an efficient calculation method for the coalition rationality coefficient based on the Monte Carlo method. Experimental results demonstrate that GT-CAM outperforms other competitive interpretation methods in visualization and quantitative analysis.

BERT-siRNA: siRNA target prediction based on BERT pre-trained interpretable model.

Silencing mRNA through siRNA is vital for RNA interference (RNAi), necessitating accurate computational methods for siRNA selection. Current approaches, relying on machine learning, often face challenges with large data requirements and intricate data preprocessing, leading to reduced accuracy. To address this challenge, we propose a BERT model-based siRNA target gene knockdown efficiency prediction method called BERT-siRNA, which consists of a pre-trained DNA-BERT module and Multilayer Perceptron module. It applies the concept of transfer learning to avoid the limitation of a small sample size and the need for extensive preprocessing processes. By fine-tuning on various siRNA datasets after pretraining on extensive genomic data using DNA-BERT to enhance predictive capabilities. Our model clearly outperforms all existing siRNA prediction models through testing on the independent public siRNA dataset. Furthermore, the model's consistent predictions of high-efficiency siRNA knockdown for SARS-CoV-2, as well as its alignment with experimental results for PDCD1, CD38, and IL6, demonstrate the reliability and stability of the model. In addition, the attention scores for all 19-nt positions in the dataset indicate that the model's attention is predominantly focused on the 5' end of the siRNA. The step-by-step visualization of the hidden layer's classification progressively clarified and explained the effective feature extraction of the MLP layer. The explainability of model by analysis the attention scores and hidden layers is also our main purpose in this work, making it more explainable and reliable for biological researchers.