The mechanism of STING autoinhibition and activation.

2',3'-cGAMP, produced by the DNA sensor cGAS, activates stimulator of interferon genes (STING) and triggers immune response during infection. Tremendous effort has been placed on unraveling the mechanism of STING activation. However, little is known about STING inhibition. Here, we found that apo-STING exhibits a bilayer with head-to-head as well as side-by-side packing, mediated by its ligand-binding domain (LBD). This type of assembly holds two endoplasmic reticulum (ER) membranes together not only to prevent STING ER exit but also to eliminate the recruitment of TBK1, representing the autoinhibited state of STING. Additionally, we obtained the filament structure of the STING/2',3'-cGAMP complex, which adopts a bent monolayer assembly mediated by LBD and transmembrane domain (TMD). The active, curved STING polymer could deform ER membrane to support its ER exit and anterograde transportation. Our data together provide a panoramic vision regarding STING autoinhibition and activation, which adds substantially to current understanding of the cGAS-STING pathway.

HimGNN: a novel hierarchical molecular graph representation learning framework for property prediction.

Accurate prediction of molecular properties is an important topic in drug discovery. Recent works have developed various representation schemes for molecular structures to capture different chemical information in molecules. The atom and motif can be viewed as hierarchical molecular structures that are widely used for learning molecular representations to predict chemical properties. Previous works have attempted to exploit both atom and motif to address the problem of information loss in single representation learning for various tasks. To further fuse such hierarchical information, the correspondence between learned chemical features from different molecular structures should be considered. Herein, we propose a novel framework for molecular property prediction, called hierarchical molecular graph neural networks (HimGNN). HimGNN learns hierarchical topology representations by applying graph neural networks on atom- and motif-based graphs. In order to boost the representational power of the motif feature, we design a Transformer-based local augmentation module to enrich motif features by introducing heterogeneous atom information in motif representation learning. Besides, we focus on the molecular hierarchical relationship and propose a simple yet effective rescaling module, called contextual self-rescaling, that adaptively recalibrates molecular representations by explicitly modelling interdependencies between atom and motif features. Extensive computational experiments demonstrate that HimGNN can achieve promising performances over state-of-the-art baselines on both classification and regression tasks in molecular property prediction.

Identification of necroptosis-associated miRNA signature for predicting prognosis and immune landscape in stomach adenocarcinoma.

PURPOSE: This study aims to identify the potential necroptosis related genes (NRGs)-associated miRNAs signature and explore the impact on the prognosis of stomach adenocarcinoma (STAD). METHODS: Employing rigorous methodologies, we utilized univariate Cox, Lasso and multivariate Cox regression analyses to develop a prognostic signature. Kaplan-Meier (K-M) and ROC curves were applied to assess the prognostic value of signature in a training group and an independent test group. Furthermore, we conducted Gene Set Enrichment Analysis (GSEA) for enrichment of tumor-related pathways. The risk score was calculated for each patient based on the expression of miRNAs which were enrolled in the signature. Patients were stratified into high- and low-risk groups. The immune cell infiltration and immunotherapy were compared between the two groups. Finally, the diagnostic potential of the miRNA was explored by RT-qPCR. RESULTS: We constructed a prognostic model based on 6 NRGs-associated miRNAs. K-M plots underscored superior survival outcomes in the low-risk group. GSEA results revealed the enrichment of several tumor-related pathways in the high-risk group. Notably, CD8+ T cells, Tregs and activated memory CD4+ T cells exhibited negative correlations with the risk score. Additionally, a few immune checkpoint genes, such as CTLA4, PD1 and PD-L1, were significantly upregulated in the low-risk group. Furthermore, the serum expression levels of all these 6 miRNAs were significantly elevated in STAD patients. CONCLUSIONS: Our study identified a robust risk score derived from a signature of 6 NRGs-associated miRNAs, demonstrating high efficacy for prognosis of STAD. These results not only contributed to our understanding of STAD pathogenesis, but also held promise for potential clinical applications, particularly in the realm of personalized immunotherapy for STAD patients.

SNX27 suppresses SARS-CoV-2 infection by inhibiting viral lysosome/late endosome entry.

After binding to its cell surface receptor angiotensin converting enzyme 2 (ACE2), severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) enters the host cell through directly fusing with plasma membrane (cell surface pathway) or undergoing endocytosis traveling to lysosome/late endosome for membrane fusion (endocytic pathway). However, the endocytic entry regulation by host cell remains elusive. Recent studies show ACE2 possesses a type I PDZ binding motif (PBM) through which it could interact with a PDZ domain-containing protein such as sorting nexin 27 (SNX27). In this study, we determined the ACE2-PBM/SNX27-PDZ complex structure, and, through a series of functional analyses, we found SNX27 plays an important role in regulating the homeostasis of ACE2 receptor. More importantly, we demonstrated SNX27, together with retromer complex (the core component of the endosomal protein sorting machinery), prevents ACE2/virus complex from entering lysosome/late endosome, resulting in decreased viral entry in cells where the endocytic pathway dominates. The ACE2/virus retrieval mediated by SNX27-retromer could be considered as a countermeasure against invasion of ACE2 receptor-using SARS coronaviruses.

META-DDIE: predicting drug-drug interaction events with few-shot learning.

Drug-drug interactions (DDIs) are one of the major concerns in pharmaceutical research, and a number of computational methods have been developed to predict whether two drugs interact or not. Recently, more attention has been paid to events caused by the DDIs, which is more useful for investigating the mechanism hidden behind the combined drug usage or adverse reactions. However, some rare events may only have few examples, hindering them from being precisely predicted. To address the above issues, we present a few-shot computational method named META-DDIE, which consists of a representation module and a comparing module, to predict DDI events. We collect drug chemical structures and DDIs from DrugBank, and categorize DDI events into hundreds of types using a standard pipeline. META-DDIE uses the structures of drugs as input and learns the interpretable representations of DDIs through the representation module. Then, the model uses the comparing module to predict whether two representations are similar, and finally predicts DDI events with few labeled examples. In the computational experiments, META-DDIE outperforms several baseline methods and especially enhances the predictive capability for rare events. Moreover, META-DDIE helps to identify the key factors that may cause DDI events and reveal the relationship among different events.

Comprehensive analysis of cuproptosis-related long noncoding RNA for predicting prognostic and diagnostic value and immune landscape in colorectal adenocarcinoma.

BACKGROUND: Cuproptosis, as a copper-induced mitochondrial cell death, has attracted extensive attention recently, especially in cancer. Although some key regulatory genes have been identified in cuproptosis, the related lncRNAs have not been further studied. Exploring the prognostic and diagnostic value of cuproptosis-related lncRNAs (CRLs) in colon adenocarcinoma and providing guidance for individualized immunotherapy for patients are of great significance. RESULTS: A total of 2003 lncRNAs were correlated with cuproptosis genes and considered as CRLs. We screened 33 survival-associated CRLs and established a prognostic signature base on 7 CRLs in the training group. The patients in the low-risk group had better outcomes in both training group (P < 0.001) and test group (P = 0.016). More exciting, our model showed good prognosis prediction in both stage I-II (P = 0.020) and stage III-IV (P = 0.001). The nomogram model could further improve the accuracy of prognosis prediction. Interestingly, glucose-related metabolic pathways, which were closely related to cuproptosis, were enriched in the low-risk group. Meanwhile, the immune infiltration scores were lower in the high-risk group. The high-risk group was more sensitive to OSI.906 and ABT.888, while low-risk group was more sensitive to Sorafenib. Three lncRNAs, FALEC, AC083967.1 and AC010997.4, were highly expressed in serum of COAD patients, and the AUC was 0.772, 0.726 and 0.714, respectively, indicating their valuable diagnostic value. CONCLUSIONS: Our research constructed a prognostic signature based on 7 CRLs and found three promising diagnostic markers for COAD patients. Our results provided a reference to the personalized immunotherapy strategies.

A systematic analysis of ARM genes revealed that GhARM144 regulates the resistance against Verticillium dahliae via interaction with GhOSM34.

Proteins of the armadillo repeat gene family play important roles in plant pathogen response. Here, 169 armadillo (ARM) genes were identified in upland cotton (Gossypium hirsutum). Phylogenetic analysis grouped these into 11 subfamilies, with conserved protein structures within each subfamily. The results signify that the expansion of the gene family occurred via whole genome duplication and dispersed duplication. Expression profiling and network analysis suggest that GhARM144 may regulate cotton resistance to Verticillium dahliae. GhARM144 was upregulated in roots by V. dahliae infection or salicylic acid treatment. This upregulation indicates a negative regulatory role of GhARM144' in the cotton immune responses, potentially by manipulating salicylic acid biosynthesis. Protein interaction studies found that GhARM144 associates with an osmotin-like protein, GhOSM34, at the plasma membrane. Silencing GhOSM34 reduced the resistance to V. dahliae, suggesting it may play a positive regulatory role. The results demonstrate that GhARM144 modulates cotton immunity through interaction with GhOSM34 and salicylic acid signalling. Further study of these proteins may yield insights into disease resistance mechanisms in cotton and other plants.

Quantitative anatomical analysis of lumbar interspaces based on 3D CT imaging: optimized segment selection for lumbar puncture in different age groups.

BACKGROUND: Optimal lumbar puncture segment selection remains controversial. This study aims to analyze anatomical differences among L3-4, L4-5, and L5-S1 segments across age groups and provide quantitative evidence for optimized selection. METHODS: 80 cases of CT images were collected with patients aged 10-80 years old. Threedimensional models containing L3-S1 vertebrae, dural sac, and nerve roots were reconstructed. Computer simulation determined the optimal puncture angles for the L3-4, L4-5, and L5-S1 segments. The effective dural sac area (ALDS), traversing nerve root area (ATNR), and area of the lumbar inter-laminar space (ALILS) were measured. Puncture efficacy ratio (ALDS/ALILS) and nerve injury risk ratio (ATNR/ALILS) were calculated. Cases were divided into four groups: A (10-20 years), B (21-40 years), C (41-60 years), and D (61-80 years). Statistical analysis was performed using SPSS. RESULTS: 1) ALDS was similar among segments; 2) ATNR was greatest at L5-S1; 3) ALILS was greatest at L5-S1; 4) Puncture efficacy ratio was highest at L3-4 and lowest at L5-S1; 5) Nerve injury risk was highest at L5-S1. In group D, L5-S1 ALDS was larger than L3-4 and L4-5. ALDS decreased after age 40. Age variations were minimal across parameters. CONCLUSION: The comprehensive analysis demonstrated L3-4 as the optimal first-choice segment for ages 10-60 years, conferring maximal efficacy and safety. L5-S1 can serve as an alternative option for ages 61-80 years when upper interspaces narrow. This study provides quantitative imaging evidence supporting age-specific, optimized lumbar puncture segment selection.

VdERG2 was involved in ergosterol biosynthesis, nutritional differentiation and virulence of Verticillium dahliae.

The ergosterol biosynthesis pathway plays an important role in model pathogenic bacteria Saccharomyces cerevisiae, but little is known about the biosynthesis of ergosterol in the pathogenic fungus Verticillium dahliae. In this study, we identified the VdERG2 gene encoding sterol C-8 isomerase from V. dahliae and investigated its function in virulence by generating gene deletion mutants (ΔVdERG2) and complemented mutants (C-ΔVdERG2). Knockout of VdERG2 reduced ergosterol content. The conidial germination rate and conidial yield of ΔVdERG2 significantly decreased and abnormal conidia were produced. In spite of VdERG2 did not affect the utilization of carbon sources by V. dahliae, but the melanin production of ΔVdERG2 was decreased in cellulose and pectin were used as the sole carbon sources. Furthermore, the ΔVdERG2 mutants produced less microsclerotia and melanin with a significant decrease in the expression of microsclerotia and melanin-related genes VaflM, Vayg1, VDH1, VdLAC, VdSCD and VT4HR. In addition, mutants ΔVdERG2 were very sensitive to congo red (CR), sodium dodecyl sulfate (SDS) and hydrogen peroxide (H2O2) stresses, indicating that VdERG2 was involved in the cell wall and oxidative stress response. The absence of VdERG2 weakened the penetration ability of mycelium on cellophane and affected the growth of mycelium. Although ΔVdERG2 could infect cotton, its pathogenicity was significantly impaired. These phenotypic defects in ΔVdERG2 could be complemented by the reintroduction of a full-length VdERG2 gene. In summary, as a single conservative secretory protein, VdERG2 played a crucial role in ergosterol biosynthesis, nutritional differentiation and virulence in V. dahliae.

Tensor decomposition with relational constraints for predicting multiple types of microRNA-disease associations.

MicroRNAs (miRNAs) play crucial roles in multifarious biological processes associated with human diseases. Identifying potential miRNA-disease associations contributes to understanding the molecular mechanisms of miRNA-related diseases. Most of the existing computational methods mainly focus on predicting whether a miRNA-disease association exists or not. However, the roles of miRNAs in diseases are prominently diverged, for instance, Genetic variants of miRNA (mir-15) may affect the expression level of miRNAs leading to B cell chronic lymphocytic leukemia, while circulating miRNAs (including mir-1246, mir-1307-3p, etc.) have potentials to detecting breast cancer in the early stage. In this paper, we aim to predict multi-type miRNA-disease associations instead of taking them as binary. To this end, we innovatively represent miRNA-disease-type triples as a tensor and introduce tensor decomposition methods to solve the prediction task. Experimental results on two widely-adopted miRNA-disease datasets: HMDD v2.0 and HMDD v3.2 show that tensor decomposition methods improve a recent baseline in a large scale (up to $38\%$ in Top-1F1). We then propose a novel method, Tensor Decomposition with Relational Constraints (TDRC), which incorporates biological features as relational constraints to further the existing tensor decomposition methods. Compared with two existing tensor decomposition methods, TDRC can produce better performance while being more efficient.

Multi-way relation-enhanced hypergraph representation learning for anti-cancer drug synergy prediction.

MOTIVATION: Drug combinations have exhibited promise in treating cancers with less toxicity and fewer adverse reactions. However, in vitro screening of synergistic drug combinations is time-consuming and labor-intensive because of the combinatorial explosion. Although a number of computational methods have been developed for predicting synergistic drug combinations, the multi-way relations between drug combinations and cell lines existing in drug synergy data have not been well exploited. RESULTS: We propose a multi-way relation-enhanced hypergraph representation learning method to predict anti-cancer drug synergy, named HypergraphSynergy. HypergraphSynergy formulates synergistic drug combinations over cancer cell lines as a hypergraph, in which drugs and cell lines are represented by nodes and synergistic drug-drug-cell line triplets are represented by hyperedges, and leverages the biochemical features of drugs and cell lines as node attributes. Then, a hypergraph neural network is designed to learn the embeddings of drugs and cell lines from the hypergraph and predict drug synergy. Moreover, the auxiliary task of reconstructing the similarity networks of drugs and cell lines is considered to enhance the generalization ability of the model. In the computational experiments, HypergraphSynergy outperforms other state-of-the-art synergy prediction methods on two benchmark datasets for both classification and regression tasks and is applicable to unseen drug combinations or cell lines. The studies revealed that the hypergraph formulation allows us to capture and explain complex multi-way relations of drug combinations and cell lines, and also provides a flexible framework to make the best use of diverse information. AVAILABILITY AND IMPLEMENTATION: The source data and codes of HypergraphSynergy can be freely downloaded from https://github.com/liuxuan666/HypergraphSynergy. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Toward drug-miRNA resistance association prediction by positional encoding graph neural network and multi-channel neural network.

Drug discovery is a costly and time-consuming process, and most drugs exert therapeutic efficacy by targeting specific proteins. However, there are a large number of proteins that are not targeted by any drug. Recently, miRNA-based therapeutics are becoming increasingly important, since miRNA can regulate the expressions of specific genes and affect a variety of human diseases. Therefore, it is of great significance to study the associations between miRNAs and drugs to enable drug discovery and disease treatment. In this work, we propose a novel method named DMR-PEG, which facilitates drug-miRNA resistance association (DMRA) prediction by leveraging positional encoding graph neural network with layer attention (LAPEG) and multi-channel neural network (MNN). LAPEG considers both the potential information in the miRNA-drug resistance heterogeneous network and the specific characteristics of entities (i.e., drugs and miRNAs) to learn favorable representations of drugs and miRNAs. And MNN models various sophisticated relations and synthesizes the predictions from different perspectives effectively. In the comprehensive experiments, DMR-PEG achieves the area under the precision-recall curve (AUPR) score of 0.2793 and the area under the receiver-operating characteristic curve (AUC) score of 0.9475, which outperforms the most state-of-the-art methods. Further experimental results show that our proposed method has good robustness and stability. The ablation study demonstrates each component in DMR-PEG is essential for drug-miRNA drug resistance association prediction. And real-world case study presents that DMR-PEG is promising for DMRA inference.

Prediction of no reflow phenomenon in percutaneous coronary intervention with optical coherence tomography and analysis of risk factors.

Objective :To investigate the predictive value of no reflow phenomenon in interventional therapy by measuring plaque quantitatively with optical coherence tomography (OCT).  Methods:196 patients with acute ST segment elevation myocardial infarction who visited the Department of Cardiology of the Second Affiliated Hospital of Zhengzhou University from January 2020 to January 2022 were selected as the study objects. According to whether there was no reflow during the operation, they were divided into no reflow group (46 cases) and normal flow group (150 cases). Systematically collect general clinical data and coronary angiography related data of patients through inpatient cases, measure fiber cap thickness and lipid core angle of diseased vascular plaque through optical coherence tomography, and analyze the relationship between fiber cap thickness and no reflow phenomenon   Results:BMI, LDL, phospholipase A, the proportion of family history of coronary heart disease, and the thrombus load in the no reflow group were higher than those in the normal flow group (P<0.05), while the thickness of the fibrous cap was lower than that in the normal flow group (P<0.05); Further multivariate logistic regression analysis showed that fiber cap thickness, phospholipase A and severe thrombosis load were independent risk factors for non reflow phenomenon (P<0.05); Further ROC curve analysis found that the thickness of fiber cap had a high predictive value for no reflow phenomenon, and the best cutoff value for no reflow was 95, AUC: 0.926 (95% CI: 0.891-0.961, P<0.001). Conclusions: Optical coherence tomography can predict the occurrence of no reflow phenomenon by measuring the fiber cap thickness quantitatively. The prediction effect is the best when the fiber cap thickness is 95.

Giant Solitary Fibrous Tumor of the Ascending Aortic Wall Causing Reversible Heart Failure: A Case Report and Review of the Literature.

A 56-year-old woman was admitted to our hospital with a 2-week history of chest tightness and fatigue, and an echocardiogram revealed a massive polyserous cavity effusion. A massive (13.5 cm maximum diameter) intrapericardial mass was discovered using computed tomography (CT) and cardiovascular magnetic resonance imaging (MRI) in the ascending aortic wall. A pericardial biopsy was performed and diagnosed as a solitary fibrous tumor (SFT). After successful mass resection, an immunohistochemical test was positive for CD34, STAT-6, CD34, and Bcl2, which indicates a giant benign solitary fibrous tumor of the ascending aortic wall. After three years of follow-up, the patient is symptom-free, and histological indications of malignancy were absent. A giant benign solitary fibrous tumor is extremely rare in the heart, especially from the ascending aorta wall, and experience with this tumor location is limited, so close follow-up at regular intervals is considered necessary. We present this case, followed by a literature review on SFTs involving the heart and management approaches.

First Report of Pestalotiopsis clavispora causing Leaf Spot on Pandanus amaryllifolius in China.

Pandanus amaryllifolius, also known as pandan, is a perennial herb, growing in Indonesia, China and the Maluku Islands (Wakte et al. 2009). It is the only plant with aromatic leaves in the Pandanaceae. It is widely used in food, medicine, cosmetics and other industries, and is also known as "Oriental Vanilla." Pandan is planted in Hainan province over 1,300 ha and is the main plant intercropped among the forest trees. From 2020, the leaf spot was surveyed for three years. Diseased leaves occurred on 30 to 80% of the surveyed plants, with an incidence of 70% and yield losses of 40%. The disease occured from mid-November to April and was most severe at low temperatures and humidity. Initial symptoms were pale green spots, that formed dark brown, nearly circular lesions. As the lesions expanded, their centers became greyish white, with yellow halos at the junction of the diseased and healthy tissue. When the humidity was high, there were small black spots scattered in the center of the lesion. Symptomatic leaf samples were collected from four different sites. The leaf surface was disinfested with 75% ethyl alcohol for 30 s and washed with sterile distilled water three times. Samples from the junction of diseased and healthy tissue (0.5 × 0.5 cm) were removed and placed on potato dextrose agar (PDA) medium containing 100 µg/mL of cefotaxime sodium and cultivated in a dark incubator at 28°C. After two days, hyphal tips from the edges of growing colonies were transferred to fresh PDA plates for further purification. Following Koch's postulates, colonies from strains were used as inoculum in pathogenicity tests. Colonies with 5 mm diameter were inoculated upside onto fresh and healthy pandan leaves via wounding method (pinpricked by sterilized needles) and non-wounding method. Sterilized PDA was used as control. All plants were setted three replicates and were incubated at 28℃ for 3 to 5 days. When symptoms on leaves similar to those in the field appeared, the fungus were reisolated The colonies formed on PDA were also consistent with the original isolate (Scandiani et al, 2003). After seven days, the colony covered the whole petri dish with white, petal-shaped growth with a slight concentric, annular bulge in the center, irregular edges, with black acervuli emerging at a later stage of colony growth. Conidia were fusiform, 18.1±1.6 × 6.4±0.3 µm, showing four septations and five cells, the middle three cells were brownish black to olivaceous, and the apical cell colorless with two to three filaments, 21.8±3.5 µm long. The caudate cell was colorless with one stalk 5.9±1.8 µm long (Zhang et al. 2021; Shu et al. 2020). According to the colony and conidia characteristics, the pathogen was initially identified as Pestalotiopsis spp. (Benjamin et al. 1961). To confirm the pathogen identity, we used the universal primers ITS1/ITS4, targeting primers EF1-728F/EF1-986R and Bt2a/Bt2b sequences (Tian et al. 2018). The sequences of the PCR products were deposited in NCBI GenBank with accession numbers OQ165166 (ITS), OQ352149 (TEF1-α) and OQ352150 (TUB2). BLAST results showed that the sequences of the ITS, TEF1-α and TUB2 genes shared 100% homology with the sequences of Pestalotiopsis clavispora. The maximum likelihood method was used in the phylogenetic analysis. The result showed that LSS112 was clustered with Pestalotiopsis clavispora with a support rate of 99%. Based on morphological and molecular characteristics, the pathogen was confirmed as Pestalotiopsis clavispora. To our knowledge, this is the first report of leaf spot of pandan caused by Pestalotiopsis clavispora in China. This research will be immediately helpful for the diagnosis and control the disease on pandan.

Identification and screening of fungicides against Piper nigrum basal Fusarium wilt disease in Hainan, China.

Fusarium wilt has occurred in the main Piper nigrum cultivation regions, which seriously affects the yield and quality of P. nigrum. To identify the pathogen of this disease, the diseased roots were collected from a demonstration base in Hainan Province. The pathogen was obtained by tissue isolation method and confirmed by pathogenicity test. Based on the morphological observation, sequence analyses of TEF1-α nuclear gene, Fusarium solani was identified as the pathogen causing P. nigrum Fusarium wilt and induced symptoms on inoculated plants, including chlorosis, necrotic spots, wilt, drying, and root rot. The experiments for the antifungal activity showed that all the 11 fungicides selected in this study showed certain inhibitory effects on the colony growth of F. solani, where 2% kasugamycin AS, 45% prochloraz EW, 25 g·L-1 fludioxonil SC and 430 g·L-1 tebuconazole SC exhibited relative higher inhibitory effects with EC50 as 0.065, 0.205, 0.395, and 0.483 mg·L-1 , respectively, and were selected to perform SEM analysis and test in seeds in vitro. The SEM analysis showed that kasugamycin, prochloraz, fludioxonil, and tebuconazole might have exerted their antifungal effect by damaging F. solani mycelia or microconidia. These preparations were applied as a seed coating of P. nigrum Reyin-1. The kasugamycin treatment was most effective in reducing the harmful impact of F. solani on the seed germination. These results presented herein provide useful guidance for the effective control of P. nigrum Fusarium wilt.

A multimodal deep learning framework for predicting drug-drug interaction events.

MOTIVATION: Drug-drug interactions (DDIs) are one of the major concerns in pharmaceutical research. Many machine learning based methods have been proposed for the DDI prediction, but most of them predict whether two drugs interact or not. The studies revealed that DDIs could cause different subsequent events, and predicting DDI-associated events is more useful for investigating the mechanism hidden behind the combined drug usage or adverse reactions. RESULTS: In this article, we collect DDIs from DrugBank database, and extract 65 categories of DDI events by dependency analysis and events trimming. We propose a multimodal deep learning framework named DDIMDL that combines diverse drug features with deep learning to build a model for predicting DDI-associated events. DDIMDL first constructs deep neural network (DNN)-based sub-models, respectively, using four types of drug features: chemical substructures, targets, enzymes and pathways, and then adopts a joint DNN framework to combine the sub-models to learn cross-modality representations of drug-drug pairs and predict DDI events. In computational experiments, DDIMDL produces high-accuracy performances and has high efficiency. Moreover, DDIMDL outperforms state-of-the-art DDI event prediction methods and baseline methods. Among all the features of drugs, the chemical substructures seem to be the most informative. With the combination of substructures, targets and enzymes, DDIMDL achieves an accuracy of 0.8852 and an area under the precision-recall curve of 0.9208. AVAILABILITY AND IMPLEMENTATION: The source code and data are available at https://github.com/YifanDengWHU/DDIMDL. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Lineage specification and pluripotency revealed by transcriptome analysis from oocyte to blastocyst in pig.

The inner cell mass (ICM) in blastocyst is the origin of all somatic and germ cells in mammals and pluripotent stem cells (PSCs) in vitro. As the conserved principles between pig and human, here we performed comprehensive single-cell RNA-seq for porcine early embryos from oocyte to early blastocyst (EB). We show the specification of the ICM and trophectoderm in morula and the molecular signature of the precursors. We demonstrate the existence of naïve pluripotency signature in morula and ICM of EB, and the specific pluripotent genes and the activity of signalling pathways highlight the characteristics of the naïve pluripotency. We observe the absence of dosage compensation with respect to X-chromosome (XC) in morula, and incomplete dosage compensation in the EB. However, the dynamics of dosage compensation may be independent of the expression of XIST induced XC inactivation. Our study describes molecular landmarks of embryogenesis in pig that will provide a better strategy for derivation of porcine PSCs and improve research in regenerative medicine.

Higenamine mitigates interleukin-1ß-induced human nucleus pulposus cell apoptosis by ROS-mediated PI3K/Akt signaling.

Intervertebral disc degeneration (IDD) is a natural problem linked to the inflammation. Higenamine exerts multiple pharmacological properties in inflammation-related disorders. Our study aimed to explore the function of higenamine on interleukin (IL)-1ß-caused apoptosis of human nucleus pulposus cells (HNPCs). Cell apoptosis was investigated by TUNEL and flow cytometry. Apoptosis-related biomarkers were determined by qRT-PCR or Western blotting. The protein in the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (Akt) signaling was measured by Western blotting. We found that higenamine showed little effect on cell apoptosis, but mitigated IL-1ß-caused apoptosis in a dose-dependent pattern. Higenamine attenuated IL-1ß-induced decrease of Bcl-2 and increase of Bax and cleaved caspase-3. Higenamine did not affect the reactive oxygen species (ROS) level and the PI3K/Akt signaling, but attenuated IL-1ß-induced ROS production and inhibition of the PI3K/Akt signaling. IL-1ß repressed the activation of the PI3K/Akt pathway, but ROS inhibition using N-acetylcysteine (NAC) rescued this pathway. The PI3K/Akt signaling suppression using LY294002 reversed the inhibitive effect of higenamine on IL-1ß-caused apoptosis, and this effect was weakened by ROS inhibition. In conclusion, higenamine attenuates IL-1ß-caused apoptosis of HNPCs via ROS-mediated PI3K/Akt pathway.

Predicting drug-drug interactions using multi-modal deep auto-encoders based network embedding and positive-unlabeled learning.

Drug-drug interactions (DDIs) are crucial for public health and patient safety, which has aroused widespread concern in academia and industry. The existing computational DDI prediction methods are mainly divided into four categories: literature extraction-based, similarity-based, matrix operations-based and network-based. A number of recent studies have revealed that integrating heterogeneous drug features is of significant importance for developing high-accuracy prediction models. Meanwhile, drugs that lack certain features could utilize other features to learn representations. However, it also brings some new challenges such as incomplete data, non-linear relations and heterogeneous properties. In this paper, we propose a multi-modal deep auto-encoders based drug representation learning method named DDI-MDAE, to predict DDIs from large-scale, noisy and sparse data. Our method aims to learn unified drug representations from multiple drug feature networks simultaneously using multi-modal deep auto-encoders. Then, we apply four operators on the learned drug embeddings to represent drug-drug pairs and adopt the random forest classifier to train models for predicting DDIs. The experimental results demonstrate the effectiveness of our proposed method for DDI prediction and significant improvement compared to other state-of-the-art benchmark methods. Moreover, we apply a specialized random forest classifier in the positive-unlabeled (PU) learning setting to enhance the prediction accuracy. Experimental results reveal that the model improved by PU learning outperforms the original method DDI-MDAE by 7.1% and 6.2% improvement in AUPR metric respectively on 3-fold cross-validation (3-CV) and 5-fold cross-validation (5-CV). And in F-measure metric, the improved model gains 10.4% and 8.4% improvement over DDI-MDAE respectively on 3-CV and 5-CV. The usefulness of DDI-MDAE is further demonstrated by case studies.