HGCLAMIR: Hypergraph contrastive learning with attention mechanism and integrated multi-view representation for predicting miRNA-disease associations.

Existing studies have shown that the abnormal expression of microRNAs (miRNAs) usually leads to the occurrence and development of human diseases. Identifying disease-related miRNAs contributes to studying the pathogenesis of diseases at the molecular level. As traditional biological experiments are time-consuming and expensive, computational methods have been used as an effective complement to infer the potential associations between miRNAs and diseases. However, most of the existing computational methods still face three main challenges: (i) learning of high-order relations; (ii) insufficient representation learning ability; (iii) importance learning and integration of multi-view embedding representation. To this end, we developed a HyperGraph Contrastive Learning with view-aware Attention Mechanism and Integrated multi-view Representation (HGCLAMIR) model to discover potential miRNA-disease associations. First, hypergraph convolutional network (HGCN) was utilized to capture high-order complex relations from hypergraphs related to miRNAs and diseases. Then, we combined HGCN with contrastive learning to improve and enhance the embedded representation learning ability of HGCN. Moreover, we introduced view-aware attention mechanism to adaptively weight the embedded representations of different views, thereby obtaining the importance of multi-view latent representations. Next, we innovatively proposed integrated representation learning to integrate the embedded representation information of multiple views for obtaining more reasonable embedding information. Finally, the integrated representation information was fed into a neural network-based matrix completion method to perform miRNA-disease association prediction. Experimental results on the cross-validation set and independent test set indicated that HGCLAMIR can achieve better prediction performance than other baseline models. Furthermore, the results of case studies and enrichment analysis further demonstrated the accuracy of HGCLAMIR and unconfirmed potential associations had biological significance.

Accurate and Convenient Lung Cancer Diagnosis through Detection of Extracellular Vesicle Membrane Proteins via Förster Resonance Energy Transfer.

Tumor-derived extracellular vesicles (EVs) are promising to monitor early stage cancer. Unfortunately, isolating and analyzing EVs from a patient's liquid biopsy are challenging. For this, we devised an EV membrane proteins detection system (EV-MPDS) based on Förster resonance energy transfer (FRET) signals between aptamer quantum dots and AIEgen dye, which eliminated the EV extraction and purification to conveniently diagnose lung cancer. In a cohort of 80 clinical samples, this system showed enhanced accuracy (100% versus 65%) and sensitivity (100% versus 55%) in cancer diagnosis as compared to the ELISA detection method. Improved accuracy of early screening (from 96.4% to 100%) was achieved by comprehensively profiling five biomarkers using a machine learning analysis system. FRET-based tumor EV-MPDS is thus an isolation-free, low-volume (1 µL), and highly accurate approach, providing the potential to aid lung cancer diagnosis and early screening.

Predicting potential microbe-disease associations based on auto-encoder and graph convolution network.

The increasing body of research has consistently demonstrated the intricate correlation between the human microbiome and human well-being. Microbes can impact the efficacy and toxicity of drugs through various pathways, as well as influence the occurrence and metastasis of tumors. In clinical practice, it is crucial to elucidate the association between microbes and diseases. Although traditional biological experiments accurately identify this association, they are time-consuming, expensive, and susceptible to experimental conditions. Consequently, conducting extensive biological experiments to screen potential microbe-disease associations becomes challenging. The computational methods can solve the above problems well, but the previous computational methods still have the problems of low utilization of node features and the prediction accuracy needs to be improved. To address this issue, we propose the DAEGCNDF model predicting potential associations between microbes and diseases. Our model calculates four similar features for each microbe and disease. These features are fused to obtain a comprehensive feature matrix representing microbes and diseases. Our model first uses the graph convolutional network module to extract low-rank features with graph information of microbes and diseases, and then uses a deep sparse Auto-Encoder to extract high-rank features of microbe-disease pairs, after which the low-rank and high-rank features are spliced to improve the utilization of node features. Finally, Deep Forest was used for microbe-disease potential relationship prediction. The experimental results show that combining low-rank and high-rank features helps to improve the model performance and Deep Forest has better classification performance than the baseline model.

Inferring circRNA-drug sensitivity associations via dual hierarchical attention networks and multiple kernel fusion.

Increasing evidence has shown that the expression of circular RNAs (circRNAs) can affect the drug sensitivity of cells and significantly influence drug efficacy. Therefore, research into the relationships between circRNAs and drugs can be of great significance in increasing the comprehension of circRNAs function, as well as contributing to the discovery of new drugs and the repurposing of existing drugs. However, it is time-consuming and costly to validate the function of circRNA with traditional medical research methods. Therefore, the development of efficient and accurate computational models that can assist in discovering the potential interactions between circRNAs and drugs is urgently needed. In this study, a novel method is proposed, called DHANMKF , that aims to predict potential circRNA-drug sensitivity interactions for further biomedical screening and validation. Firstly, multimodal networks were constructed by DHANMKF using multiple sources of information on circRNAs and drugs. Secondly, comprehensive intra-type and inter-type node representations were learned using bi-typed multi-relational heterogeneous graphs, which are attention-based encoders utilizing a hierarchical process. Thirdly, the multi-kernel fusion method was used to fuse intra-type embedding and inter-type embedding. Finally, the Dual Laplacian Regularized Least Squares method (DLapRLS) was used to predict the potential circRNA-drug sensitivity associations using the combined kernel in circRNA and drug spaces. Compared with the other methods, DHANMKF obtained the highest AUC value on two datasets. Code is available at https://github.com/cuntjx/DHANMKF .

Schumannella soli sp. nov., a novel actinomycete isolated from mangrove soil by in situ cultivation.

A novel actinobacterial strain, designated 10F1D-1T, was isolated from soil sample collected from Futian mangrove nature reserve, China using of the in situ cultivation technique. Preliminary analysis based on the 16S rRNA gene sequence revealed that strain 10F1D-1T was the member of genus Schumannella with sharing highest sequence similarity (99.7%) to Schumannella luteola DSM 23141T. Phylogenetic analyses based on 16S rRNA gene sequences and core proteome consistently exhibited that strain 10F1D-1T formed a monophyletic clade with Schumannella luteola DSM 23141T. Comparative genomic analyses clearly separated strain 10F1D-1T from the only species of the genus Schumannella based on average nucleotide identity (ANI), average amino acid identity (AAI) and digital DNA-DNA hybridization (dDDH) values below the thresholds for species delineation. The genome of strain 10F1D-1T contains the biosynthetic gene clusters for osmoprotectants to adapt to the salt environment of mangrove. Strain 10F1D-1T also contains the biosynthetic gene clusters for bioactive compounds as secondary metabolites. On the basis of the polyphasic analysis, strain 10F1D-1T is considered to represent a novel species of the genus Schumannella, for which the name Schumannella soli sp. nov. (type strain 10F1D-1T = CGMCC1.16699T = JCM 33146T) is proposed.

Amnibacterium endophyticum sp. nov., an endophytic actinobacterium isolated from Aegiceras corniculatum.

A Gram-stain-positive, aerobic, non-motile, non-spore-forming and short-rod-shaped actinobacterium, designated strain 1T4Z-3T, was isolated from a piece of surface-sterilized branch of Aegiceras corniculatum collected from the Cotai Ecological Zones in Macao, China. Comparative 16S rRNA gene sequence analysis showed that strain 1T4Z-3T was clearly affiliated to the genus Amnibacterium and exhibited 97.9 % gene sequence similarity to Amnibacterium kyonggiense JCM 16463T, 97.3 % gene sequence similarity to Amnibacterium soli JCM 19015T and less than 96.4 % gene sequence similarities to other genera of the family Microbacteriaceae. Strain 1T4Z-3T had L-2,4-diaminobutyric acid as the diagnostic cell-wall diamino acid. The major fatty acids (>10 % of total fatty acids) were iso-C16 : 0 (46.6 %) and anteiso-C15 : 0 (27.3 %). The predominant menaquinones of strain 1T4Z-3T were MK-11 (81.4 %) and MK-12 (14.1 %). The polar lipids comprised diphosphatidylglycerol, phosphatidylglycerol, six unidentified glycolipids, four unidentified phospholipids and two unidentified lipids. The DNA G+C content of strain 1T4Z-3T was 71.4 mol%. Based on the phylogenetic, phenotypic and chemotaxonomic features, strain 1T4Z-3T is considered to represent a novel species of the genus Amnibacterium, for which the name Amnibacterium endophyticum sp. nov. is proposed. The type strain of Amnibacterium endophyticum is 1T4Z-3T (=KCTC 39983T=CGMCC 1.16066T).

Mangrovicella endophytica gen. nov., sp. nov., a new member of the family Aurantimonadaceae isolated from Aegiceras corniculatum.

A Gram-negative, aerobic, motile and short-rod-shaped bacterium, designated strain 5T4P-12-1T, was isolated from a piece of surface-sterilized bark of Aegiceras corniculatum collected from Cotai Ecological Zones in Macao, China and tested by a polyphasic approach to clarify its taxonomic position. Strain 5T4P-12-1T grew optimally with 0-1 % (w/v) NaCl at 30 °C and at pH 7.0-8.0. The 16S rRNA gene sequence of strain 5T4P-12-1T had the highest similarity (96.7 %) to Aureimonas altamirensis DSM 21988T. Phylogenic analysis based on 16S rRNA gene sequences and coding sequences of 98 protein clusters showed that the strain represented a novel genus of the family Aurantimonadaceae. The predominant quinone system of strain 5T4P-12-1T was ubiquinone 10. The polar lipid profile contained diphosphatidylglycerol, phosphatidylglycerol, phosphatidylcholine, phosphatidylethanolamine, phosphatidylmethylethanolamine, an unidentified aminophospholipid, three unidentified aminolipids, three unidentified phospholipids and three unidentified lipids. The major fatty acids (>10 % of total fatty acids) were C18 : 1ω7c (55.4 %) and C18 : 1 2-OH (15.6 %). The DNA G+C content of strain 5T4P-12-1T was 66.5 mol%. Based on the phylogenic, phenotypic and chemotaxonomic features, strain 5T4P-12-1T is considered to represent a novel species of a new genus in the family Aurantimonadaceae, for which the name Mangrovicella endophytica gen. nov., sp. nov. is proposed. The type strain is 5T4P-12-1T (=KCTC 62053T=CGMCC 1.16279 T).

Self-supervised learning-based Multi-Scale feature Fusion Network for survival analysis from whole slide images.

Understanding prognosis and mortality is critical for evaluating the treatment plan of patients. Advances in digital pathology and deep learning techniques have made it practical to perform survival analysis in whole slide images (WSIs). Current methods are usually based on a multi-stage framework which includes patch sampling, feature extraction and prediction. However, the random patch sampling strategy is highly unstable and prone to sampling non-ROI. Feature extraction typically relies on hand-crafted features or convolutional neural networks (CNNs) pre-trained on ImageNet, while the artificial error or domain gaps may affect the survival prediction performance. Besides, the limited information representation of local sampling patches will create a bottleneck limitation on the effectiveness of prediction. To address the above challenges, we propose a novel patch sampling strategy based on image information entropy and construct a Multi-Scale feature Fusion Network (MSFN) based on self-supervised feature extractor. Specifically, we adopt image information entropy as a criterion to select representative sampling patches, thereby avoiding the noise interference caused by random to blank regions. Meanwhile, we pretrain the feature extractor utilizing self-supervised learning mechanism to improve the efficiency of feature extraction. Furthermore, a global-local feature fusion prediction network based on the attention mechanism is constructed to improve the survival prediction effect of WSIs with comprehensive multi-scale information representation. The proposed method is validated by adequate experiments and achieves competitive results on both of the most popular WSIs survival analysis datasets, TCGA-GBM and TCGA-LUSC. Code and trained models are made available at: https://github.com/Mercuriiio/MSFN.

Inferring human miRNA-disease associations via multiple kernel fusion on GCNII.

Increasing evidence shows that the occurrence of human complex diseases is closely related to the mutation and abnormal expression of microRNAs(miRNAs). MiRNAs have complex and fine regulatory mechanisms, which makes it a promising target for drug discovery and disease diagnosis. Therefore, predicting the potential miRNA-disease associations has practical significance. In this paper, we proposed an miRNA-disease association predicting method based on multiple kernel fusion on Graph Convolutional Network via Initial residual and Identity mapping (GCNII), called MKFGCNII. Firstly, we built a heterogeneous network of miRNAs and diseases to extract multi-layer features via GCNII. Secondly, multiple kernel fusion method was applied to weight fusion of embeddings at each layer. Finally, Dual Laplacian Regularized Least Squares was used to predict new miRNA-disease associations by the combined kernel in miRNA and disease spaces. Compared with the other methods, MKFGCNII obtained the highest AUC value of 0.9631. Code is available at https://github.com/cuntjx/bioInfo.

Metagenomic Analysis Reveals Microbial Community Structure and Metabolic Potential for Nitrogen Acquisition in the Oligotrophic Surface Water of the Indian Ocean.

Despite being the world's third largest ocean, the Indian Ocean is one of the least studied and understood with respect to microbial diversity as well as biogeochemical and ecological functions. In this study, we investigated the microbial community and its metabolic potential for nitrogen (N) acquisition in the oligotrophic surface waters of the Indian Ocean using a metagenomic approach. Proteobacteria and Cyanobacteria dominated the microbial community with an average 37.85 and 23.56% of relative abundance, respectively, followed by Bacteroidetes (3.73%), Actinobacteria (1.69%), Firmicutes (0.76%), Verrucomicrobia (0.36%), and Planctomycetes (0.31%). Overall, only 24.3% of functional genes were common among all sampling stations indicating a high level of gene diversity. However, the presence of 82.6% common KEGG Orthology (KOs) in all samples showed high functional redundancy across the Indian Ocean. Temperature, phosphate, silicate and pH were important environmental factors regulating the microbial distribution in the Indian Ocean. The cyanobacterial genus Prochlorococcus was abundant with an average 17.4% of relative abundance in the surface waters, and while 54 Prochlorococcus genomes were detected, 53 were grouped mainly within HLII clade. In total, 179 of 234 Prochlorococcus sequences extracted from the global ocean dataset were clustered into HL clades and exhibited less divergence, but 55 sequences of LL clades presented more divergence exhibiting different branch length. The genes encoding enzymes related to ammonia metabolism, such as urease, glutamate dehydrogenase, ammonia transporter, and nitrilase presented higher abundances than the genes involved in inorganic N assimilation in both microbial community and metagenomic Prochlorococcus population. Furthermore, genes associated with dissimilatory nitrate reduction, denitrification, nitrogen fixation, nitrification and anammox were absent in metagenome Prochlorococcus population, i.e., nitrogenase and nitrate reductase. Notably, the de novo biosynthesis pathways of six different amino acids were incomplete in the metagenomic Prochlorococcus population and Prochlorococcus genomes, suggesting compensatory uptake of these amino acids from the environment. These results reveal the features of the taxonomic and functional structure of the Indian Ocean microbiome and their adaptive strategies to ambient N deficiency in the oligotrophic ocean.

Deciphering the Microbial Taxonomy and Functionality of Two Diverse Mangrove Ecosystems and Their Potential Abilities To Produce Bioactive Compounds.

Mangroves, as important and special ecosystems, create unique ecological environments for examining the microbial gene capacity and potential for producing bioactive compounds. However, little is known about the biogeochemical implications of microbiomes in mangrove ecosystems, especially the variations between pristine and anthropogenic mangroves. To elucidate this, we investigated the microbial taxonomic and functional shifts of the mangrove microbiomes and their potential for bioactive compounds in two different coastal mangrove ecosystems in southern China. A gene catalogue, including 87 million unique genes, was constructed, based on deep shotgun metagenomic sequencing. Differentially enriched bacterial and archaeal taxa between pristine mangroves (Guangxi) and anthropogenic mangroves (Shenzhen) were found. The Nitrospira and ammonia-oxidizing archaea, specifically, were more abundant in Shenzhen mangroves, while sulfate-reducing bacteria and methanogens were more abundant in Guangxi mangroves. The results of functional analysis were consistent with the taxonomic results, indicating that the Shenzhen mangrove microbiome has a higher abundance of genes involved in nitrogen metabolism while the Guangxi mangrove microbiome has a higher capacity for sulfur metabolism and methanogenesis. Biosynthetic gene clusters were identified in the metagenome data and in hundreds of de novo reconstructed nonredundant microbial genomes, respectively. Notably, we found different biosynthetic potential in different taxa, and we identified three high quality and novel Acidobacteria genomes with a large number of BGCs. In total, 67,278 unique genes were annotated with antibiotic resistance, indicating the prevalence and persistence in multidrug-resistant genes in the mangrove microbiome.IMPORTANCE This study comprehensively described the taxonomy and functionality of mangrove microbiomes, including their capacity for secondary metabolite biosynthesis and their ability to resist antibiotics. The microbial taxonomic and functional characteristics differed between geographical locations, corresponding to the environmental condition of two diverse mangrove regions. A large number of microbial biosynthetic gene clusters encoding novel bioactivities were found, and this can serve as a valuable resource to guide novel bioactive compound discovery for potential clinical uses.

Aeromicrobium endophyticum sp. nov., an endophytic actinobacterium isolated from reed (Phragmites australis).

A Gram-staining-positive, motile and short-rod-shaped actinobacterium designated 9W16Y-2T was isolated from surface-sterilized leaves of reed (Phragmites australis) collected from Taklamakan Desert in Xinjiang Uygur Autonomous Region, China. Colonies were pale greenish yellow, circular, smooth, and convex. The 16S rRNA gene sequence of strain 9W16Y-2T exhibited highest sequence similarities with Aeromicrobium camelliae CGMCC 1.12942T (99.0%) and Aeromicrobium erythreum NRRL B-3381T (97.2%). Phylogenetic analyses based on 16S rRNA gene sequences and single-copy phylogenetic marker genes (pMGs) showed that strain 9W16Y-2T belonged to the genus Aeromicrobium and formed a monophyletic clade with Aeromicrobium camelliae CGMCC 1.12942T. Furthermore, average nucleotide identity (ANI) and DNA-DNA hybridization (DDH) clearly separated strain 9W16Y-2T from the other species of the genus Aeromicrobium with values below the thresholds for species delineation. The G+C content of the genomic DNA is 68.9 mol%. The diagnostic diamino acid of the cell-wall peptidoglycan was LL-diaminopimelic acid. The predominant menaquinone was MK-9(H4). The major fatty acids (> 10% of the total fatty acids) were C18:0 10-methyl (TBSA) (28.2%), C16:0 (21.0%), C16:0 2-OH (20.8%) and C18:1ω9c (12.8%). The polar lipid profile comprised diphosphatidylglycerol, phosphatidylglycerol, phosphatidylcholine, phosphatidylinositol, an unidentified aminophospholipid and an unidentified lipid. Based on the phylogenic, phenotypic and chemotaxonomic features, strain 9W16Y-2T represents a novel species of the genus Aeromicrobium, for which the name Aeromicrobium endophyticum sp. nov. is proposed. The type strain is 9W16Y-2T (= CGMCC 1.13876T = JCM 33141T).

Labedella phragmitis sp. nov. and Labedella populi sp. nov., two endophytic actinobacteria isolated from plants in the Taklamakan Desert and emended description of the genus Labedella.

Two novel strains, designated 11W25H-1T and 8H24J-4-2T, were isolated from surface-sterilized plant tissues collected from the Taklamakan Desert in the Xinjiang Uygur Autonomous Region, China. The strains were characterized by a polyphasic approach in order to clarify their taxonomic positions. They were Gram-stain positive, aerobic, non-motile, non-spore-forming and rod-shaped. The 16S rRNA gene sequences of the strains showed highest similarities with Labedella gwakjiensis KCTC 19176T (99.2% and 98.9%, respectively) and Labedella endophytica CPCC 203961T (98.9% and 99.0%, respectively). The sequence similarity between strains 11W25H-1T and 8H24J-4-2T was 99.4%. Phylogenetic analyses based on 16S rRNA gene sequences and single-copy phylogenetic marker genes (pMGs) showed that the two strains belonged to the genus Labedella and formed a separate cluster from the closest species L. gwakjiensis KCTC 19176T and L. endophytica CPCC 203961T. Genomic analyses, including average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH), clearly separated the strains from each other and from the other species of the genus Labedella with values below the thresholds for species delineation. The two strains showed chemotaxonomic characteristics and phenotypic properties in agreement with the description of the genus Labedella and also confirmed the differentiation from the closest species. The data demonstrated that strains 11W25H-1T and 8H24J-4-2T represented two novel species of the genus Labedella, for which the names Labedella phragmitis sp. nov. (type strain 11W25H-1T=JCM 33144T=CGMCC 1.16700T) and Labedella populi sp. nov. (type strain 8H24J-4-2T=JCM 33143T=CGMCC 1.16697T) are proposed.

Deciphering the Composition and Functional Profile of the Microbial Communities in Chinese Moutai Liquor Starters.

Moutai is a world-famous traditional Chinese liquor with complex taste and aroma, which are considered to be strongly influenced by the quality of fermentation starters (Daqu). However, the role of microbial communities in the starters has not been fully understood. In this study, we revealed the microbial composition of 185 Moutai starter samples, covering three different types of starters across immature and mature phases, and functional gene composition of mature starter microbiome. Our results showed that microbial composition patterns of immature starters varied, but they eventually were similar and steady when they became mature starters, after half-year storage and subsequent mixing. To help identify two types of immature starters, we selected seven operational taxonomic unit (OTU) markers by leave-one-out cross validation (LOOCV) and an OTU classified as Saccharopolyspora was the most decisive one. For mature starters, we identified a total of 16 core OTUs, one of which annotated as Bacillus was found positively associated with saccharifying power. We also identified the functional gene and microbial composition in starch and cellulose hydrolysis pathways. Microbes with higher abundances of alpha-glucosidase, alpha-amylase, and glucoamylase probably contributed to high saccharifying power. Overall, this study reveals the features of Moutai starter microbial communities in different phases and improves understanding of the relationships between microbiota and functional properties of the starters.