Model fusion for predicting unconventional proteins secreted by exosomes using deep learning

Unconventional secretory proteins (USPs) are vital for cell-to-cell communication and are necessary for proper physiological processes. Unlike classical proteins that follow the conventional secretory pathway via the Golgi apparatus, these proteins are released using unconventional pathways. The primary modes of secretion for USPs are exosomes and ectosomes, which originate from the endoplasmic reticulum. Accurate and rapid identification of exosome-mediated secretory proteins is crucial for gaining valuable insights into the regulation of non-classical protein secretion and intercellular communication, as well as for the advancement of novel therapeutic approaches. Although computational methods based on amino acid sequence prediction exist for predicting unconventional proteins secreted by exosomes (UPSEs), they suffer from significant limitations in terms of algorithmic accuracy. In this study, we propose a novel approach to predict UPSEs by combining multiple deep learning models that incorporate both protein sequences and evolutionary information. Our approach utilizes a convolutional neural network (CNN) to extract protein sequence information, while various densely connected neural networks (DNNs) are employed to capture evolutionary conservation patterns.By combining six distinct deep learning models, we have created a superior framework that surpasses previous approaches, achieving an ACC score of 77.46% and an MCC score of 0.5406 on an independent test dataset.

Identifying diagnostic indicators for type 2 diabetes mellitus from physical examination using interpretable machine learning approach

Background

Identification of patients at risk for type 2 diabetes mellitus (T2DM) can not only prevent complications and reduce suffering but also ease the health care burden. While routine physical examination can provide useful information for diagnosis, manual exploration of routine physical examination records is not feasible due to the high prevalence of T2DM.

Objectives

We aim to build interpretable machine learning models for T2DM diagnosis and uncover important diagnostic indicators from physical examination, including age- and sex-related indicators.

Methods

In this study, we present three weighted diversity density (WDD)-based algorithms for T2DM screening that use physical examination indicators, the algorithms are highly transparent and interpretable, two of which are missing value tolerant algorithms.

Patients

Regarding the dataset, we collected 43 physical examination indicator data from 11,071 cases of T2DM patients and 126,622 healthy controls at the Affiliated Hospital of Southwest Medical University. After data processing, we used a data matrix containing 16004 EHRs and 43 clinical indicators for modelling.

Results

The indicators were ranked according to their model weights, and the top 25% of indicators were found to be directly or indirectly related to T2DM. We further investigated the clinical characteristics of different age and sex groups, and found that the algorithms can detect relevant indicators specific to these groups. The algorithms performed well in T2DM screening, with the highest area under the receiver operating characteristic curve (AUC) reaching 0.9185.

Conclusion

This work utilized the interpretable WDD-based algorithms to construct T2DM diagnostic models based on physical examination indicators. By modeling data grouped by age and sex, we identified several predictive markers related to age and sex, uncovering characteristic differences among various groups of T2DM patients.

EnsembleDL-ATG: Identifying autophagy proteins by integrating their sequence and evolutionary information using an ensemble deep learning framework

Autophagy is a primary mechanism for maintaining cellular homeostasis. The synergistic actions of autophagy-related (ATG) proteins strictly regulate the whole autophagic process. Therefore, accurate identification of ATGs is a first and critical step to reveal the molecular mechanism underlying the regulation of autophagy. Current computational methods can predict ATGs from primary protein sequences, but owing to the limitations of algorithms, significant room for improvement still exists. In this research, we propose EnsembleDL-ATG, an ensemble deep learning framework that aggregates multiple deep learning models to predict ATGs from protein sequence and evolutionary information. We first evaluated the performance of individual networks for various feature descriptors to identify the most promising models. Then, we explored all possible combinations of independent models to select the most effective ensemble architecture. The final framework was built and maintained by an organization of four different deep learning models. Experimental results show that our proposed method achieves a prediction accuracy of 94.5 % and MCC of 0.890, which are nearly 4 % and 0.08 higher than ATGPred-FL, respectively. Overall, EnsembleDL-ATG is the first ATG machine learning predictor based on ensemble deep learning. The benchmark data and code utilized in this study can be accessed for free at https://github.com/jingry/autoBioSeqpy/tree/2.0/examples/EnsembleDL-ATG.

Fast and Efficient Design of Deep Neural Networks for Predicting N7-Methylguanosine Sites Using autoBioSeqpy

N⁷-Methylguanosine (m⁷G) is a crucial post-transcriptional RNA modification that plays a pivotal role in regulating gene expression. Accurately identifying m⁷G sites is a fundamental step in understanding the biological functions and regulatory mechanisms associated with this modification. While whole-genome sequencing is the gold standard for RNA modification site detection, it is a time-consuming, expensive, and intricate process. Recently, computational approaches, especially deep learning (DL) techniques, have gained popularity in achieving this objective. Convolutional neural networks and recurrent neural networks are examples of DL algorithms that have emerged as versatile tools for modeling biological sequence data. However, developing an efficient network architecture with superior performance remains a challenging task, requiring significant expertise, time, and effort. To address this, we previously introduced a tool called autoBioSeqpy, which streamlines the design and implementation of DL networks for biological sequence classification. In this study, we utilized autoBioSeqpy to develop, train, evaluate, and fine-tune sequence-level DL models for predicting m⁷G sites. We provided detailed descriptions of these models, along with a step-by-step guide on their execution. The same methodology can be applied to other systems dealing with similar biological questions. The benchmark data and code utilized in this study can be accessed for free at http://github.com/jingry/autoBioSeeqpy/tree/2.0/examples/m7G.

Evaluation and development of deep neural networks for RNA 5-Methyluridine classifications using autoBioSeqpy

Post-transcriptionally RNA modifications, also known as the epitranscriptome, play crucial roles in the regulation of gene expression during development. Recently, deep learning (DL) has been employed for RNA modification site prediction and has shown promising results. However, due to the lack of relevant studies, it is unclear which DL architecture is best suited for some pyrimidine modifications, such as 5-methyluridine (m⁵U). To fill this knowledge gap, we first performed a comparative evaluation of various commonly used DL models for epigenetic studies with the help of autoBioSeqpy. We identified optimal architectural variations for m⁵U site classification, optimizing the layer depth and neuron width. Second, we used this knowledge to develop Deepm5U, an improved convolutional-recurrent neural network that accurately predicts m⁵U sites from RNA sequences. We successfully applied Deepm5U to transcriptomewide m⁵U profiling data across different sequencing technologies and cell types. Third, we showed that the techniques for interpreting deep neural networks, including LayerUMAP and DeepSHAP, can provide important insights into the internal operation and behavior of models. Overall, we offered practical guidance for the development, benchmark, and analysis of deep learning models when designing new algorithms for RNA modifications.

Negative refraction in hyperbolic hetero-bicrystals

We visualized negative refraction of phonon polaritons, which occurs at the interface between two natural crystals. The polaritons-hybrids of infrared photons and lattice vibrations-form collimated rays that display negative refraction when passing through a planar interface between the two hyperbolic van der Waals materials: molybdenum oxide (MoO₃) and isotopically pure hexagonal boron nitride (h¹¹BN). At a special frequency ω₀, these rays can circulate along closed diamond-shaped trajectories. We have shown that polariton eigenmodes display regions of both positive and negative dispersion interrupted by multiple gaps that result from polaritonic-level repulsion and strong coupling.

Multi-model predictive analysis of RNA solvent accessibility based on modified residual attention mechanism

Predicting RNA solvent accessibility using only primary sequence data can be regarded as sequence-based prediction work. Currently, the established studies for sequence-based RNA solvent accessibility prediction are limited due to the available number of datasets and black box prediction. To improve these issues, we first expanded the available RNA structures and then developed a sequence-based model using modified attention layers with different receptive fields to conform to the stem-loop structure of RNA chains. We measured the improvement with an extended dataset and further explored the model's interpretability by analysing the model structures, attention values and hyperparameters. Finally, we found that the developed model regarded the pieces of a sequence as templates during the training process. This work will be helpful for researchers who would like to build RNA attribute prediction models using deep learning in the future.

Systematic Analysis and Accurate Identification of DNA N4-Methylcytosine Sites by Deep Learning

DNA N⁴-methylcytosine (4mC) is a pivotal epigenetic modification that plays an essential role in DNA replication, repair, expression and differentiation. To gain insight into the biological functions of 4mC, it is critical to identify their modification sites in the genomics. Recently, deep learning has become increasingly popular in recent years and frequently employed for the 4mC site identification. However, a systematic analysis of how to build predictive models using deep learning techniques is still lacking. In this work, we first summarized all existing deep learning-based predictors and systematically analyzed their models, features and datasets, etc. Then, using a typical standard dataset with three species (A. thaliana, C. elegans, and D. melanogaster), we assessed the contribution of different model architectures, encoding methods and the attention mechanism in establishing a deep learning-based model for the 4mC site prediction. After a series of optimizations, convolutional-recurrent neural network architecture using the one-hot encoding and attention mechanism achieved the best overall prediction performance. Extensive comparison experiments were conducted based on the same dataset. This work will be helpful for researchers who would like to build the 4mC prediction models using deep learning in the future.

The applications of deep learning algorithms on in silico druggable proteins identification

•

We developed the first deep learning-based druggable protein classifier for fast and accurate identification of potential druggable proteins.

•

Experimental results on a standard dataset demonstrate that the prediction performance of deep learning model is comparable to those of existing methods.

•

We visualized the representations of druggable proteins learned by deep learning models, which helps us understand how they work.

•

Our analysis reconfirms that the attention mechanism is especially useful for explaining deep learning models.

Introduction

The top priority in drug development is to identify novel and effective drug targets. In vitro assays are frequently used for this purpose; however, traditional experimental approaches are insufficient for large-scale exploration of novel drug targets, as they are expensive, time-consuming and laborious. Therefore, computational methods have emerged in recent decades as an alternative to aid experimental drug discovery studies by developing sophisticated predictive models to estimate unknown drugs/compounds and their targets. The recent success of deep learning (DL) techniques in machine learning and artificial intelligence has further attracted a great deal of attention in the biomedicine field, including computational drug discovery.

Objectives

This study focuses on the practical applications of deep learning algorithms for predicting druggable proteins and proposes a powerful predictor for fast and accurate identification of potential drug targets.

Methods

Using a gold-standard dataset, we explored several typical protein features and different deep learning algorithms and evaluated their performance in a comprehensive way. We provide an overview of the entire experimental process, including protein features and descriptors, neural network architectures, libraries and toolkits for deep learning modelling, performance evaluation metrics, model interpretation and visualization.

Results

Experimental results show that the hybrid model (architecture: CNN-RNN (BiLSTM) + DNN; feature: dictionary encoding + DC_TC_CTD) performed better than the other models on the benchmark dataset. This hybrid model was able to achieve 90.0% accuracy and 0.800 MCC on the test dataset and 84.8% and 0.703 on a nonredundant independent test dataset, which is comparable to those of existing methods.

Conclusion

We developed the first deep learning-based classifier for fast and accurate identification of potential druggable proteins. We hope that this study will be helpful for future researchers who would like to use deep learning techniques to develop relevant predictive models.

DeepT3 2.0: improving type III secreted effector predictions by an integrative deep learning framework

Type III secretion systems (T3SSs) are bacterial membrane-embedded nanomachines that allow a number of humans, plant and animal pathogens to inject virulence factors directly into the cytoplasm of eukaryotic cells. Export of effectors through T3SSs is critical for motility and virulence of most Gram-negative pathogens. Current computational methods can predict type III secreted effectors (T3SEs) from amino acid sequences, but due to algorithmic constraints, reliable and large-scale prediction of T3SEs in Gram-negative bacteria remains a challenge. Here, we present DeepT3 2.0 (http://advintbioinforlab.com/deept3/), a novel web server that integrates different deep learning models for genome-wide predicting T3SEs from a bacterium of interest. DeepT3 2.0 combines various deep learning architectures including convolutional, recurrent, convolutional-recurrent and multilayer neural networks to learn N-terminal representations of proteins specifically for T3SE prediction. Outcomes from the different models are processed and integrated for discriminating T3SEs and non-T3SEs. Because it leverages diverse models and an integrative deep learning framework, DeepT3 2.0 outperforms existing methods in validation datasets. In addition, the features learned from networks are analyzed and visualized to explain how models make their predictions. We propose DeepT3 2.0 as an integrated and accurate tool for the discovery of T3SEs.

Fizeau drag in graphene plasmonics

Dragging of light by moving media was predicted by Fresnel¹ and verified by Fizeau's celebrated experiments² with flowing water. This momentous discovery is among the experimental cornerstones of Einstein's special relativity theory and is well understood^3,4 in the context of relativistic kinematics. By contrast, experiments on dragging photons by an electron flow in solids are riddled with inconsistencies and have so far eluded agreement with the theory^5-7. Here we report on the electron flow dragging surface plasmon polaritons^8,9 (SPPs): hybrid quasiparticles of infrared photons and electrons in graphene. The drag is visualized directly through infrared nano-imaging of propagating plasmonic waves in the presence of a high-density current. The polaritons in graphene shorten their wavelength when propagating against the drifting carriers. Unlike the Fizeau effect for light, the SPP drag by electrical currents defies explanation by simple kinematics and is linked to the nonlinear electrodynamics of Dirac electrons in graphene. The observed plasmonic Fizeau drag enables breaking of time-reversal symmetry and reciprocity¹⁰ at infrared frequencies without resorting to magnetic fields^11,12 or chiral optical pumping^13,14. The Fizeau drag also provides a tool with which to study interactions and nonequilibrium effects in electron liquids.

Prediction of disease-associated functional variants in noncoding regions through a comprehensive analysis by integrating datasets and features

One of the greatest challenges in human genetics is deciphering the link between functional variants in noncoding sequences and the pathophysiology of complex diseases. To address this issue, many methods have been developed to sort functional single-nucleotide variants (SNVs) for neutral SNVs in noncoding regions. In this study, we integrated well-established features and commonly used datasets and merged them into large-scale datasets based on a random forest model, which yielded promising performance and outperformed some cutting-edge approaches. Our analyses of feature importance and data coverage also provide certain clues for future research in enhancing the prediction of functional noncoding SNVs.

MiR-808 inhibits cardiomyocyte apoptosis and expressions of caspase-3 and caspase-9 in rats with myocardial infarction by regulating TGF-β1 signaling pathway

OBJECTIVE

To investigate the effects of micro ribonucleic acid (miR)-808 on cardiomyocyte apoptosis and expressions of caspase-3 and caspase-9 in rats with myocardial infarction (MI) by regulating the transforming growth factor-β1 (TGF-β1) signaling pathway.

MATERIALS AND METHODS

A total of 24 specific pathogen-free female Sprague-Dawley rats were enrolled and randomly divided into normal group, model group, and miR-808 group, 8 rats in each group. In the model group and miR-808 group, MI model was prepared by ligation of the left anterior descending coronary artery in the rats. The miR-808 group was transfected with miR-808 lentivirus after the model was established. After one week of intervention, the expression of TGF-β1 was detected by reverse transcription-polymerase chain reaction (RT-PCR). The cardiac function of rats was determined by echocardiography. The myocardium of rats was observed by Masson staining. The cardiomyocyte apoptosis of rats was examined by TdT-mediated dUTP-biotin nick end labeling (TUNEL) method. The expression levels of caspase-3 and caspase-9 were detected by Western blotting.

RESULTS

The expression of TGF-β1 mRNA was higher in the model group than that in the normal group (p<0.05), but compared with that in the model group, it was lower in the miR-808 group. The myocardial function and cardiomyocyte survival rate in the miR-808 group was better and higher than those in the model group (p<0.05). The expression levels of caspase-3 and caspase-9 in the miR-808 group were lower than those in the model group (p<0.05).

CONCLUSIONS

MiR-808 can inhibit cardiomyocyte apoptosis in rats with MI by down-regulating TGF-β1 expression and inhibiting the expressions of caspase-3 and caspase-9.

STRAP reduces endoplasmic reticulum stress and apoptosis in cardiomyocytes and attenuates myocardial ischemia-reperfusion injury by activating PI3K/PDK1/Akt signaling pathway

OBJECTIVE

Myocardial ischemia-reperfusion injury (MIRI) is a common problem in heart-related diseases. The aim of this study was to explore the protective effects of STRAP on cardiomyocytes in the MIRI process and its mechanisms.

MATERIALS AND METHODS

We used SD rats to construct a MIRI model and increased the expression of STRAP in myocardial tissue by Entranster to detect the effect of STRAP on rat myocardial tissue. In addition, we cultured rat cardiomyocyte cell line H9c2 cells and constructed a hypoxia-reoxygenation model to detect the protective effect of STRAP on H9c2 cells. LY294002, an inhibitor of the PI3K/PDK1/Akt signaling pathway, was used to validate the mechanism by which STRAP protects cardiomyocytes.

RESULTS

Overexpression of STRAP significantly reduced the activity of MDA in myocardial tissue and increased the activity of SOD. STRAP also substantially lowered CK and LDH levels in rat serum and increased Na+-K+-ATPase and Ca2+-Mg2+-ATPase activity. In addition, overexpression of STRAP considerably reduced endoplasmic reticulum stress (ERS) and apoptosis levels in H9c2 cells. However, LY294002 attenuated the protective effect of STRAP on cardiomyocytes.

CONCLUSIONS

STRAP reduces ERS and apoptosis in cardiomyocytes by activating the PI3K/PDK1/Akt signaling pathway, thereby reducing myocardial MIRI.

DeepT3_4: A Hybrid Deep Neural Network Model for the Distinction Between Bacterial Type III and IV Secreted Effectors

Gram-negative bacteria can deliver secreted proteins (also known as secreted effectors) directly into host cells through type III secretion system (T3SS), type IV secretion system (T4SS), and type VI secretion system (T6SS) and cause various diseases. These secreted effectors are heavily involved in the interactions between bacteria and host cells, so their identification is crucial for the discovery and development of novel anti-bacterial drugs. It is currently challenging to accurately distinguish type III secreted effectors (T3SEs) and type IV secreted effectors (T4SEs) because neither T3SEs nor T4SEs contain N-terminal signal peptides, and some of these effectors have similar evolutionary conserved profiles and sequence motifs. To address this challenge, we develop a deep learning (DL) approach called DeepT3_4 to correctly classify T3SEs and T4SEs. We generate amino-acid character dictionary and sequence-based features extracted from effector proteins and subsequently implement these features into a hybrid model that integrates recurrent neural networks (RNNs) and deep neural networks (DNNs). After training the model, the hybrid neural network classifies secreted effectors into two different classes with an accuracy, F-value, and recall of over 80.0%. Our approach stands for the first DL approach for the classification of T3SEs and T4SEs, providing a promising supplementary tool for further secretome studies.

DeepACP: A Novel Computational Approach for Accurate Identification of Anticancer Peptides by Deep Learning Algorithm

Cancer is one of the most dangerous diseases to human health. The accurate prediction of anticancer peptides (ACPs) would be valuable for the development and design of novel anticancer agents. Current deep neural network models have obtained state-of-the-art prediction accuracy for the ACP classification task. However, based on existing studies, it remains unclear which deep learning architecture achieves the best performance. Thus, in this study, we first present a systematic exploration of three important deep learning architectures: convolutional, recurrent, and convolutional-recurrent networks for distinguishing ACPs from non-ACPs. We find that the recurrent neural network with bidirectional long short-term memory cells is superior to other architectures. By utilizing the proposed model, we implement a sequence-based deep learning tool (DeepACP) to accurately predict the likelihood of a peptide exhibiting anticancer activity. The results indicate that DeepACP outperforms several existing methods and can be used as an effective tool for the prediction of anticancer peptides. Furthermore, we visualize and understand the deep learning model. We hope that our strategy can be extended to identify other types of peptides and may provide more assistance to the development of proteomics and new drugs.

IL-6 knockout ameliorates myocardial remodeling after myocardial infarction by regulating activation of M2 macrophages and fibroblast cells

OBJECTIVE

To investigate the effects of interleukin-6 (IL-6) gene knockout on myocardial remodeling after myocardial infarction (MI) in mice and the potential mechanism, to provide certain references for the prevention and treatment of MI in clinic.

MATERIALS AND METHODS

A total of 40 male C57 mice were divided into two groups, namely Sham group (n=20) and MI group (n=20), using a random number table. Another 20 mice with IL-6 gene knockout were enrolled into the MI + IL-6 KO group. The MI model was established by means of ligating the left anterior descending coronary artery of the mice. 28 d later, the survival status of the three groups of mice was recorded. In addition, the cardiac functions of each group of mice, including two-dimensional echocardiography, ejection fraction (EF%) and fractional shortening (FS%), were measured. The cross-sectional area and pathological change of the myocardial cells in cardiac tissues of each group of mice were detected via hematoxylin and eosin (H&E) staining. Immunohistochemistry was applied to determine the expression of tumor necrosis factor-alpha (TNF-α) in each group of mouse cardiac tissues. Moreover, immunofluorescent staining was utilized to measure the content of M2 macrophages in each group of mouse cardiac tissues.

RESULTS

The 28-d survival rate of the mice with IL-6 gene knockout was remarkably higher than that of the wild-type mice (p<0.05). Furthermore, the cardiac functions of the mice in the MI + IL-6 KO group were superior to those in the MI group, with markedly improved FS% and EF% (p<0.05). According to the H&E staining results, the cross-sectional areas of the heart and myocardial cells were decreased notably in MI + IL-6 KO group compared with those in the MI group (p<0.05). The immunohistochemical staining results showed that IL-6 knockout could lower the MI-induced high expression of TNF-α (p<0.05), and Masson's trichrome staining indicated that IL-6 knockout could also repress the degree of cardiac fibrosis. Moreover, it was discovered through immunofluorescent staining that the mice in the MI + IL-6 KO group had markedly elevated content of M2 macrophages in cardiac tissues than those in the MI group (p<0.05).

CONCLUSIONS

Inhibiting IL-6 gene expression can prominently ameliorate the MI-induced myocardial remodeling, whose mechanism is possibly associated with the activation of M2 macrophages and reduced collagen production in fibroblast cells.

Downregulated miRNA-26a-5p induces the apoptosis of endothelial cells in coronary heart disease by inhibiting PI3K/AKT pathway

OBJECTIVE

Multiple microRNAs (miRNAs) are abnormally expressed in endothelial cells during the occurrence of coronary artery disease (CAD). Previous researches have demonstrated that miRNA-26a-5p participates in regulating the proliferation of vascular smooth muscle cells and angiogenesis. The aim of this study was to clarify the role of miRNA-26a-5p in regulating cellular performances of endothelial cells in the progression of CAD.

PATIENTS AND METHODS

In vivo CAD model was successfully established by feeding high-fat diet in 8-week-old female ApoE/LDLR-/- mice. CAD mice were administered with miRNA-26a-5p NC or miRNA-26a-5p inhibitor, respectively. Meanwhile, coronary endothelial cells were isolated from CAD mice and normal controls. Relative levels of miRNA-26a-5p, the gene of phosphate and tension homology deleted on chromosome ten (PTEN) and vascular endothelial growth factor (VEGF) in CAD patients and coronary endothelial cells isolated from CAD mice were examined. The regulatory effect of miRNA-26a-5p on atherosclerosis-related genes in primary endothelial cells and HUVECs were detected as well. Moreover, the viability and apoptosis of primary endothelial cells with miRNA-26a-5p knockdown were assessed by cell counting kit-8 (CCK-8) assay and flow cytometry, respectively. Dual-luciferase reporter gene assay was conducted to identify the relationship between miRNA-26a-5p and PTEN. Furthermore, the regulatory role of miRNA-26a-5p in phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) pathway was examined in endothelial cells.

RESULTS

MiRNA-26a-5p and VEGF were significantly downregulated in CAD patients and primary endothelial cells isolated from CAD mice. However, PTEN was significantly upregulated. CAD mice administrated with miRNA-26a-5p inhibitor exhibited remarkably upregulated ET-1, TxA2, and ANG II, as well as downregulated eNOS and PGI2. Conversely, transfection of miRNA-26a-5p mimics in HUVECs obtained the opposite trends. PTEN was identified as the direct target gene of miRNA-26a-5p. Moreover, significantly reduced viability and enhanced apoptotic rate were observed in endothelial cells isolated from CAD mice administrated with miRNA-26a-5p inhibitor. In addition, the protein level of p-AKT in endothelial cells with miRNA-26a-5p knockdown was significantly down-regulated.

CONCLUSIONS

MiRNA-26a-5p influences the proliferative and apoptotic abilities of endothelial cells isolated from CAD mice by targeting PTEN to activate PI3K/AKT pathway.

autoBioSeqpy: A Deep Learning Tool for the Classification of Biological Sequences

Deep learning has proven to be a powerful method with applications in various fields including image, language, and biomedical data. Thanks to the libraries and toolkits such as TensorFlow, PyTorch, and Keras, researchers can use different deep learning architectures and data sets for rapid modeling. However, the available implementations of neural networks using these toolkits are usually designed for a specific research and are difficult to transfer to other work. Here, we present autoBioSeqpy, a tool that uses deep learning for biological sequence classification. The advantage of this tool is its simplicity. Users only need to prepare the input data set and then use a command line interface. Then, autoBioSeqpy automatically executes a series of customizable steps including text reading, parameter initialization, sequence encoding, model loading, training, and evaluation. In addition, the tool provides various ready-to-apply and adapt model templates to improve the usability of these networks. We introduce the application of autoBioSeqpy on three biological sequence problems: the prediction of type III secreted proteins, protein subcellular localization, and CRISPR/Cas9 sgRNA activity. autoBioSeqpy is freely available with examples at https://github.com/jingry/autoBioSeqpy.

Narrowing the Gap Between In Vitro and In Vivo Genetic Profiles by Deconvoluting Toxicogenomic Data In Silico

Toxicogenomics (TGx) is a powerful method to evaluate toxicity and is widely used in both in vivo and in vitro assays. For in vivo TGx, reduction, refinement, and replacement represent the unremitting pursuit of live-animal tests, but in vitro assays, as alternatives, usually demonstrate poor correlation with real in vivo assays. In living subjects, in addition to drug effects, inner-environmental reactions also affect genetic variation, and these two factors are further jointly reflected in gene abundance. Thus, finding a strategy to factorize inner-environmental factor from in vivo assays based on gene expression levels and to further utilize in vitro data to better simulate in vivo data is needed. We proposed a strategy based on post-modified non-negative matrix factorization, which can estimate the gene expression profiles and contents of major factors in samples. The applicability of the strategy was first verified, and the strategy was then utilized to simulate in vivo data by correcting in vitro data. The similarities between real in vivo data and simulated data (single-dose 0.72, repeat-doses 0.75) were higher than those observed when directly comparing real in vivo data with in vitro data (single-dose 0.56, repeat-doses 0.70). Moreover, by keeping environment-related factor, a simulation can always be generated by using in vitro data to provide potential substitutions for in vivo TGx and to reduce the launch of live-animal tests.

[The predictive value of epicardial adipose tissue and inflammatory factors for in-stent restenosis]

Objective: To investigate the predictive value of epicardial adipose tissue volume (EATV) and inflammatory factors on in-stent restenosis (ISR) after percutaneous coronary implantation (PCI) in patients with coronary heart disease (CAD). Methods: A total of 407 patients with CAD who were treated with drug-eluting stents in TEDA international cardiovascular disease hospital were enrolled from November 2016 to October 2017. Levels of inflammatory cytokines such as high sensitive c-reactive protein (Hs-CRP), interleukin-6 (IL-6), and tumor necrosis factor (TNF-α) were detected. EATV was measured preoperatively by multi-sliced CT. Patients were divided into ISR group (n=52) and N-ISR group (n=355) according to ISR occurred within 1 year after procedure. The relationship between EATV and inflammatory factors and ISR after PCI was analyzed. Results: The differences between ISR group (n=52) and N-ISR group (n=355) were statistically significant in terms of diabetes history, IL-6, TNF-α, EATV ((150±36) cm(3)vs(120±40) cm(3),P=0.001)), bifurcation lesions, stent length and Gensini score (P<0.05). Multivariate Logistic regression analysis results showed that diabetes history,bifurcation lesions, TNF-α, EATV, and Gensini score were risk factors for in-stent restenosis.The area under the ROC curve (AUC) of EATV, TNF-α, and IL-6 in patients with CAD after PCI was 0.712, 0.752 and 0.675 (95%CI 0.648-0.776, 0.686-0.819, 0.584-0.766, respectively, all P<0.001), with a sensitivity of 86.5%, 67.3% and 69.2%, a specificity of 53.8%, 74.4% and 70.1% and a cut-off value of 116.61 cm(3),138.40 µg/L and 126.4 µg/L, respectively. Conclusion: EATV, TNF-α, and IL-6 have certain predictive values for in-stent restenosis, and can be used as clinical indicators to predict in-stent restenosis.

[Relationship between epicardial adipose tissue and clinical prognosis of patients with coronary heart disease after percutaneous coronary intervention]

Objective: To further evaluate the clinical value of epicardial adipose tissue volume (EATV) in predicting the prognosis of coronary heart disease (CHD) after percutaneous coronary intervention (PCI). Methods: From July 2013 to July 2016 in TEDA International Cardiovascular Disease Hospital, a total of 474 patients diagnosed with CHD were included in this study.According to the result of EATV, patients were divided into three groups, group A (EATV≤75 ml), group B (75 ml<EATV<150 ml), and group C (EATV≥150 ml). Then the level of body mass index (BMI), hypersensitive c-reactive protein (hs-CRP), interleukin (IL)-6 and tumor necrosis factor (TNF)-α were tested for all the three groups.All the patients were followed up for 1 year for major adverse cardiovascular events (MACE). The clinical value of EATV in predicting the occurrence of MACE events was evaluated. Results: The BMI, level of hs-CRP, TNF-α in group B were higher than group A, group C were significantly higher than group B, with statistically significant difference across all the comparisons (P<0.05). Spearman correlation analysis showed EATV was positively correlated with hs-CRP, IL-6, TNF-α (r=0.675-0.700, P<0.01). The incidence of MACE in the three groups were 8.50% in group C, 5.26% in group B, 3.13% in group A, and the differences were all significant (P<0.01). ROC curve showed the cut-off value of EATV level was 120.39 ml to predict MACE (area under cure: 0.751, 95%CI: 0.634-0.868, P<0.01), and the sensitivity was 72.7%, the specificity was 61%.EATV>120.39 ml can be used as an independent risk factor for predicting the occurrence of MACE. Conclusion: The level of EATV is closely related to the occurrence of MACE events, and EATV>120.39 ml is an independent risk factor for MACE in patients with CHD after PCI.

[Effects of epicardial adipose tissue and inflammatory factors on left ventricular diastolic function in patients with coronary heart disease]

Objective: To investigate the effects ofepicardial adipose tissue volume (EATV) and inflammatory factors on left ventricular diastolic function in patients with coronary heart disease(CHD). Methods: The clinical data of patients with coronary heart disease receiving coronary artery intervention therapy from January 2014 to October 2015 in TEDA international cardiovascular hospital were preoperatively collected.We measured the indexes of EATV and left ventricular diastolic function. Results: The difference of age (F=7.76, P=0.01), IL-6 (F=14.34, P<0.01), Hs-CRP (F=4.08, P=0.04), adiponect-in (F=4.50, P=0.04) and EATV (F=71.29, P<0.01) between the diastolicdysfunction group (n=156) and the normal group (n=76) was statistically significant.Multivariate logistic regression analysis showed that EATV was a risk factor for left ventricular diastolic dysfunction in patients with coronary artery disease (P<0.05), OR=1.05, 95%CI (1.03-1.06). The AUC value of EATV in the diagnosis of left ventriculardiastolic function in patients with coronary heart disease was 0.79, 95%CI (0.73-0.85) P<0.01. Conclusions: EATV can be used as an independent risk factor for left ventricular diastolic dysfunction.It has some non-invasive diagnosis and predictive value, and it can be used as a new therapeutic target.

CIRBP protects H9C2 cells against myocardial ischemia through inhibition of NF-κB pathway

Myocardial ischemia is a major cause of death and remains a disease with extremely deficient clinical therapies and a major problem worldwide. Cold inducible RNA-binding protein (CIRBP) is reported to be involved in multiple pathological processes, including myocardial ischemia. However, the molecular mechanisms of myocardial ischemia remain elusive. Here, we first overexpressed CIRBP by transfection of pc-CIRBP (pcDNA3.1 containing coding sequenced for CIRBP) and silenced CIRBP by transfection of small interfering RNA targeting CIRBP (siCIRBP). pcDNA3.1 and the negative control of siCIRBP (siNC) were transfected into H9C2 cells to act as controls. We then constructed a cell model of myocardial ischemia through culturing cells in serum-free medium with hypoxia in H9C2 cells. Subsequently, AlamarBlue assay, flow cytometry and western blot analysis were used, respectively, to assess cell viability, reactive oxygen species (ROS) level and apoptosis, and expression levels of IκBα, p65 and Bcl-3. We demonstrated that CIRBP overexpression promoted cell proliferation (P<0.001), inhibited cell apoptosis (P<0.05), reduced ROS level (P<0.001), down-regulated phosphorylated levels of IκBα and p65 (P<0.01 or P<0.001), and up-regulated expression of Bcl-3 (P<0.001) in H9C2 cells with myocardial ischemia. The influence of CIRBP knockdown yielded opposite results. Our study revealed that CIRBP could protect H9C2 cells against myocardial ischemia through inhibition of NF-κB pathway.

[Situation of long-term use of oral anticoagulation among atrial fibrillation patients with stroke in different level hospital]

OBJECTIVE

To investigate the current situation, time trends and factors associated with long-term use of oral anticoagulation (OAC) among atrial fibrillation (AF) patients with ischemic stroke.

METHODS

We used the dataset from the CAFR (Chinese Atrial Fibrillation Registry), a prospective, multicenter, hospital-based registry study involving 20 tertiary and 12 nontertiary hospitals in Beijing. In brief, 380 consecutive AF patients with following ischemic stroke were enrolled from 2003 to 2014.Patients with valvular AF, radiofrequency catheter ablation history or contraindications of OAC were excluded. We divided the patients into two groups according to hospital level, and investigated the rate of OAC use and its change over time in patients who had indication, the factors including patient characteristics and hospital level associated with OAC use were also analyzed.

RESULTS

Overall oral anticoagulation use rate was 27.71%, which dropped to 22.11% and 15.26% at 6 months and 12 months, respectively.A total of 298 participates were enrolled from tertiary hospitals (78.42%), and 82 were enrolled from nontertiary hospitals. The status of OAC use in tertiary hospitals was better than nontertiary hospitals (32.66% vs 7.32%, P<0.001). Multivariable analysis showed better oral anticoagulation use was independently associated with higher-level hospitals (odds ratio 1.785, 95% confidence interval 1.026-3.106, P=0.040), and history of heart failure (odds ratio 2.247, 95% confidence interval 1.235-4.090, P=0.008).

CONCLUSIONS

These data indicates oral anticoagulation use has improved in atrial fibrillation patients with stroke in Beijing. The use of anticoagulation among the patients from tertiary hospitals is significantly better than those from nontertiary hospitals, and the history of heart failure may have effect on the use of oral anticoagulation.

The incidence and economic burden of injuries in Jiangxi, China

OBJECTIVES

This study estimated the incidence, direct medical and non-medical costs, and productivity losses due to morbidity and mortality across multiple strata for injuries that occurred in Jiangxi, China.

STUDY DESIGN

Cross-sectional study.

METHODS

Data came from the Jiangxi injury survey, a provincially-representative, population-based sample of 100,010 households. The major economic costs of injuries were divided into direct costs and indirect costs. Direct costs encompass medical costs and direct non-medical costs. Indirect costs refer to the productivity losses due to injury-related morbidity and mortality.

RESULTS

In 2005, about one of 18 residents in Jiangxi, China, experienced an injury. Overall, fall, animal bite, and road traffic crash (RTC) injuries accounted for more than 66% of all injuries, while fall, RTC, drowning, and self-harm injuries accounted for 80% of fatal injuries. Average cost per case for a fatal injury was 163,389 RMB ($20,171) for lost productivity and 2800 RMB ($346) in direct medical & non-medical costs. A non-fatal injury resulting in hospitalisation or permanent disability on average caused 5221 RMB ($643) in direct costs and 18,437 RMB ($2276) in lost productivity and, an additional loss of three school days. A non-hospitalised non-fatal injury on average caused 303 ($37) RMB in direct costs and 491 RMB ($61) in lost productivity and, an additional loss of 0.5 school days.

CONCLUSIONS

The unequivocal evidence of the substantial health and financial burden of injuries indicates to Chinese policy makers that more research and efforts are needed to find efficacious and cost-effective interventions targeting injury.

A facile strategy applied to simultaneous qualitative-detection on multiple components of mixture samples: a joint study of infrared spectroscopy and multi-label algorithms on PBX explosives

We combined infrared spectroscopy with multi-label algorithms to propose a facile yet efficient strategy to realize simultaneous qualitative-detection on multiple components of mixture explosives without pre-separation.

Comparative analysis of oncogenes identified by microarray and RNA-sequencing as biomarkers for clinical prognosis

Aims: Although RNA-sequencing has been widely used to identify the differentially expressed genes (DEGs) as biomarkers to guide the therapeutic treatment, it is necessary to investigate the concordance of DEGs identified by microarray and RNA-sequencing for the clinical prognosis. Material & methods: By using The Cancer Genome Atlas data sets, we thoroughly investigated the concordance of DEGs identified from microarray and RNA-sequencing data and their molecular functions. Results: The DEGs identified by both technologies averaged ˜98.6% overlap. The cancer-related gene sets were significantly enriched with the DEGs and consistent between two technologies. Conclusions: The highly consistency of DEGs in their regulation directionality and molecular functions indicated the good reproducibility between microarray and RNA-sequencing in identifying potential oncogenes for clinical prognosis.

Domain position prediction based on sequence information by using fuzzy mean operator

The prediction of protein domain region is an advantageous process on the study of protein structure and function. In this study, we proposed a new method, which is composed of fuzzy mean operator and region division, to predict the particular positions of domains in a target protein based on its sequence. The whole sequence is aligned and scored by using fuzzy mean operator, and the final determination of domain region position is realized by region division. A published benchmark is used for the comparison with previous researches. In addition, we generate two extra datasets to examine the stability of this method. Finally, the prediction accuracy of independent test dataset achieved by our method was up to 84.13%. We wish that this method could be useful for related researches.

Characteristic wavenumbers of Raman spectra reveal the molecular mechanisms of oral leukoplakia and can help to improve the performance of diagnostic models

An effective method for diagnosing various grades of oral leukoplakia with dysplasia.

Genetic diversity in European Pisum germplasm collections

The distinctness of, and overlap between, pea genotypes held in several Pisum germplasm collections has been used to determine their relatedness and to test previous ideas about the genetic diversity of Pisum. Our characterisation of genetic diversity among 4,538 Pisum accessions held in 7 European Genebanks has identified sources of novel genetic variation, and both reinforces and refines previous interpretations of the overall structure of genetic diversity in Pisum. Molecular marker analysis was based upon the presence/absence of polymorphism of retrotransposon insertions scored by a high-throughput microarray and SSAP approaches. We conclude that the diversity of Pisum constitutes a broad continuum, with graded differentiation into sub-populations which display various degrees of distinctness. The most distinct genetic groups correspond to the named taxa while the cultivars and landraces of Pisum sativum can be divided into two broad types, one of which is strongly enriched for modern cultivars. The addition of germplasm sets from six European Genebanks, chosen to represent high diversity, to a single collection previously studied with these markers resulted in modest additions to the overall diversity observed, suggesting that the great majority of the total genetic diversity collected for the Pisum genus has now been described. Two interesting sources of novel genetic variation have been identified. Finally, we have proposed reference sets of core accessions with a range of sample sizes to represent Pisum diversity for the future study and exploitation by researchers and breeders.

Identification of microRNA-target interaction in APRIL-knockdown colorectal cancer cells

MicroRNAs (miRNAs) regulate mammalian gene expression by targeting mRNAs and have key roles in several cellular processes, including differentiation, development, apoptosis and cancer pathomechanisms. Our previous studies have confirmed that a proliferation-inducing ligand (APRIL) gene is overexpressed in colorectal cancer (CRC) tumors and SW480 cells. To study the potential mechanisms of APRIL gene in the occurrence and development of the CRC, herein, we investigated whether APRIL-knockdown had the inhibitory effect on the growth of SW480 cells and had the simultaneous expression changes of miRNAs and mRNAs by microarrays. Our results suggest that siRNA-APRIL can effectively inhibit the growth of SW480 cells in vitro and in vivo and several miRNAs via specific pathways might be involved in regulating the phenotype of loss-of-function in APRIL-knockdown SW480 cells. Thus, our study highlights the possible mechanisms of miRNA-target regulating the function of APRIL gene in CRC cells, moreover, siRNA-APRIL holds great promise as a novel gene therapy approach for APRIL- positive CRC treatment.

Receptor for advanced glycation end products (RAGE) soluble form (sRAGE): a new biomarker for lung cancer

Receptor for advanced glycation end products (RAGE) may be involved in the pathogenesis of cancer progression. Pathological effects mediated via RAGE are physiologically inhibited by soluble RAGE (sRAGE). The aim of this study was to identify sRAGE and RAGE expression profile in lung cancer patients. An ELISA method was used to quantify serum sRAGE in 45 individuals. Additionally, surgical specimens of 28 lung cancer patients were also included for RAGE expression by immunohistochemistry. Serum sRAGE was significantly decreased in lung cancer patients compared with controls (vs. healthy donors, P=0.034; vs. pulmonary tuberculosis patients, P=0.010). Lower sRAGE concentration was negative correlated with lymph node involvement (N0 vs. N1-2, P=0.028). Down regulation of membranous and cytoplasmic expression for RAGE was also lower in lung cancer tissue than in nearby normal lung tissue. Correlation with serum sRAGE concentration and RAGE expression in lung cancer tissue was existed by CV values. The results indicate that serum sRAGE levels are decreased during lung cancer progression and could reflect decreased RAGE expression in tissue. Serum sRAGE may serve as an effective and convenient diagnostic biomarker for lung cancer.

KEYWORDS

sRAGE, serum, RAGE, tissue, lung cancer.

[Purification and properties of glutamate dehydrogenase from Pseudomonas pseudoalcaligenes]

Glutamate dehydrogenase was purified from the crude extract of Pseudomonas pseudoal-caligenes. The enzyme had a molecular weight of 290,000 and was composed of six subunits with identical molecular weight of 47,000. The enzyme was highly specific for NADP(H) and the substrates. The biochemical properties such as kinetic parameters and heat stability were also examined. The purified GDH showed considerable loss of activity upon freezing.

[Clinical studies on changes in sexual hormones and estrogen receptor in patients with gynecomastia]

Levels of serum testosterone (T), estradiol (E2), and estrogen receptor (ER) in the tissue were detected by radioimmunoassay and immunohistochemical method (ABC technique) in 46 patients with gynecomastia. The results showed that rise of E2 and E2/T is a whole-body etiologic factor of hypertrophy of male breast, and the positive rate of ER is an important local etiologic factor of gynecomastia. E2 and ER have aosynergism effect on onset and development of gynecomastia and cause the ductal cell hyperplasia of male breast. It should be alerted whether it can cause male breast cancer.