Predicting the potential human lncRNA-miRNA interactions based on graph convolution network with conditional random field

A comprehensive survey on deep learning-based identification and predicting the interaction mechanism of long non-coding RNAs

Long noncoding RNAs (lncRNAs) have been discovered to be extensively involved in eukaryotic epigenetic, transcriptional, and post-transcriptional regulatory processes with the advancements in sequencing technology and genomics research. Therefore, they play crucial roles in the body's normal physiology and various disease outcomes. Presently, numerous unknown lncRNA sequencing data require exploration. Establishing deep learning-based prediction models for lncRNAs provides valuable insights for researchers, substantially reducing time and costs associated with trial and error and facilitating the disease-relevant lncRNA identification for prognosis analysis and targeted drug development as the era of artificial intelligence progresses. However, most lncRNA-related researchers lack awareness of the latest advancements in deep learning models and model selection and application in functional research on lncRNAs. Thus, we elucidate the concept of deep learning models, explore several prevalent deep learning algorithms and their data preferences, conduct a comprehensive review of recent literature studies with exemplary predictive performance over the past 5 years in conjunction with diverse prediction functions, critically analyze and discuss the merits and limitations of current deep learning models and solutions, while also proposing prospects based on cutting-edge advancements in lncRNA research.

Identifying vital nodes for yeast network by dynamic network entropy

BACKGROUND

The progress of the cell cycle of yeast involves the regulatory relationships between genes and the interactions proteins. However, it is still obscure which type of protein plays a decisive role in regulation and how to identify the vital nodes in the regulatory network. To elucidate the sensitive node or gene in the progression of yeast, here, we select 8 crucial regulatory factors from the yeast cell cycle to decipher a specific network and propose a simple mixed K2 algorithm to identify effectively the sensitive nodes and genes in the evolution of yeast.

RESULTS

Considering the multivariate of cell cycle data, we first utilize the K2 algorithm limited to the stationary interval for the time series segmentation to measure the scores for refining the specific network. After that, we employ the network entropy to effectively screen the obtained specific network, and simulate the gene expression data by a normal distribution approximation and the screened specific network by the partial least squares method. We can conclude that the robustness of the specific network screened by network entropy is better than that of the specific network with the determined relationship by comparing the obtained specific network with the determined relationship. Finally, we can determine that the node CDH1 has the highest score in the specific network through a sensitivity score calculated by network entropy implying the gene CDH1 is the most sensitive regulatory factor.

CONCLUSIONS

It is clearly of great potential value to reconstruct and visualize gene regulatory networks according to gene databases for life activities. Here, we present an available algorithm to achieve the network reconstruction by measuring the network entropy and identifying the vital nodes in the specific nodes. The results indicate that inhibiting or enhancing the expression of CDH1 can maximize the inhibition or enhancement of the yeast cell cycle. Although our algorithm is simple, it is also the first step in deciphering the profound mystery of gene regulation.

Predictive value of inflammatory prognostic index for contrast-induced nephropathy in patients undergoing coronary angiography and/or percutaneous coronary intervention

The purpose of this study was to investigate the relationship between Inflammatory Prognostic Index (IPI) levels and Contrast-Induced Nephropathy (CIN) risk and postoperative clinical outcomes in patients undergoing coronary angiography (CAG) and/or percutaneous coronary intervention (PCI). A total of 3,340 consecutive patients who underwent CAG and/or PCI between May 2017 and December 2022 were enrolled in this study. Based on their baseline IPI levels, patients were categorized into four groups. Clinical characteristics and postoperative outcomes were compared among these groups. In-hospital outcomes focused on CIN risk, repeated revascularization, major bleeding, and major adverse cardiovascular events (MACEs), while the long-term outcome examined the all-cause readmission rate. Quartile analysis found a significant link between IPI levels and CIN risk, notably in the highest quartile (P < 0.001). Even after adjusting for baseline factors, this association remained significant, with an adjusted Odds Ratio (aOR) of 2.33 (95%CI 1.50-3.64; P = 0.001). Notably, baseline IPI level emerged as an independent predictor of severe arrhythmia, with aOR of 0.50 (95%CI 0.35-0.69; P < 0.001), particularly driven by the highest quartile. Furthermore, a significant correlation between IPI and acute myocardial infarction was observed (P < 0.001), which remained significant post-adjustment. For patients undergoing CAG and/or PCI, baseline IPI levels can independently predict clinical prognosis. As a comprehensive inflammation indicator, IPI effectively identifies high-risk patients post-procedure. This study underscores IPI's potential to assist medical professionals in making more precise clinical decisions, ultimately reducing mortality and readmission rates linked to cardiovascular disease (CVD).

SCTC: inference of developmental potential from single-cell transcriptional complexity

Inferring the developmental potential of single cells from scRNA-Seq data and reconstructing the pseudo-temporal path of cell development are fundamental but challenging tasks in single-cell analysis. Although single-cell transcriptional diversity (SCTD) measured by the number of expressed genes per cell has been widely used as a hallmark of developmental potential, it may lead to incorrect estimation of differentiation states in some cases where gene expression does not decrease monotonously during the development process. In this study, we propose a novel metric called single-cell transcriptional complexity (SCTC), which draws on insights from the economic complexity theory and takes into account the sophisticated structure information of scRNA-Seq count matrix. We show that SCTC characterizes developmental potential more accurately than SCTD, especially in the early stages of development where cells typically have lower diversity but higher complexity than those in the later stages. Based on the SCTC, we provide an unsupervised method for accurate, robust, and transferable inference of single-cell pseudotime. Our findings suggest that the complexity emerging from the interplay between cells and genes determines the developmental potential, providing new insights into the understanding of biological development from the perspective of complexity theory.

In silico analysis unveiling potential biomarkers in gallbladder carcinogenesis

Gallbladder cancer (GBC) is a rare but very aggressive most common digestive tract cancer with a high mortality rate due to delayed diagnosis at the advanced stage. Moreover, GBC progression shows asymptomatic characteristics making it impossible to detect at an early stage. In these circumstances, conventional therapy like surgery, chemotherapy, and radiotherapy becomes refractive. However, few studies reported some molecular markers like KRAS (Kirsten Rat Sarcoma) mutation, upregulation of HER2/neu, EGFR (Epidermal Growth Factor Receptor), and microRNAs in GBC. However, the absence of some specific early diagnostic and prognostic markers is the biggest hurdle for the therapy of GBC to date. The present study has been designed to identify some specific molecular markers for precise diagnosis, and prognosis, for successful treatment of the GBC. By In Silico a network-centric analysis of two microarray datasets; (GSE202479) and (GSE13222) from the Gene Expression Omnibus (GEO) database, shows 50 differentially expressed genes (DEGs) associated with GBC. Further network analysis revealed that 12 genes are highly interconnected based on the highest MCODE (Molecular Complex Detection) value, among all three genes; TRIP13 (Thyroid Receptor Interacting Protein), NEK2 (Never in Mitosis gene-A related Kinase 2), and TPX2 (Targeting Protein for Xklp2) having highest network interaction with transcription factors and miRNA suggesting critically associated with GBC. Further survival analysis data corroborate the association of these genes; TRIP13, NEK2, and TPX2 with GBC. Thus, TRIP13, NEK2, and TPX2 genes are significantly correlated with a greater risk of mortality, transforming them from mere biomarkers of the GBC for early detections and may emerge as prognostic markers for treatment.

Hybrid transformer-CNN model for accurate prediction of peptide hemolytic potential

Hemolysis is a crucial factor in various biomedical and pharmaceutical contexts, driving our interest in developing advanced computational techniques for precise prediction. Our proposed approach takes advantage of the unique capabilities of convolutional neural networks (CNNs) and transformers to detect complex patterns inherent in the data. The integration of CNN and transformers' attention mechanisms allows for the extraction of relevant information, leading to accurate predictions of hemolytic potential. The proposed method was trained on three distinct data sets of peptide sequences known as recurrent neural network-hemolytic (RNN-Hem), Hlppredfuse, and Combined. Our computational results demonstrated the superior efficacy of our models compared to existing methods. The proposed approach demonstrated impressive Matthews correlation coefficients of 0.5962, 0.9111, and 0.7788 respectively, indicating its effectiveness in predicting hemolytic activity. With its potential to guide experimental efforts in peptide design and drug development, this method holds great promise for practical applications. Integrating CNNs and transformers proves to be a powerful tool in the fields of bioinformatics and therapeutic research, highlighting their potential to drive advancement in this area.

To explore the prognostic characteristics of colon cancer based on tertiary lymphoid structure-related genes and reveal the characteristics of tumor microenvironment and drug prediction

In order to construct a prognostic evaluation model of TLS features in COAD and better realize personalized precision medicine in COAD. Colon adenocarcinoma (COAD) is a common malignant tumor of the digestive system. At present, there is no effective prognostic marker to predict the prognosis of patients. Tertiary lymphoid structure (TLS) affects cancer progression by regulating immune microenvironment. Mining COAD biomarkers based on TLS-related genes helps to improve the prognosis of patients. In order to construct a prognostic evaluation model of TLS features in COAD and better realize personalized precision medicine in COAD. The mRNA expression data and clinical information of COAD and adjacent tissues were downloaded from the Cancer Genome Atlas database. The differentially expressed TLS-related genes of COAD relative to adjacent tissues were obtained by differential analysis. TLS gene co-expression analysis was used to mine genes highly related to TLS, and the intersection of the two was used to obtain candidate genes. Univariate, LASSO, and multivariate Cox regression analysis were performed on candidate genes to screen prognostic markers to construct a risk assessment model. The differences of immune characteristics were evaluated by ESTIMATE, ssGSEA and CIBERSORT in high and low risk groups of prognostic model. The difference of genomic mutation between groups was evaluated by tumor mutation burden score. Screening small molecule drugs through the GDSC library. Finally, a nomogram was drawn to evaluate the clinical value of the prognostic model. Seven TLS-related genes ADAM8, SLC6A1, PAXX, RIMKLB, PTH1R, CD1B, and MMP10 were screened to construct a prognostic model. Survival analysis showed that patients in the high-risk group had significantly lower overall survival rates. Immune microenvironment analysis showed that patients in the high-risk group had higher immune indicators, indicating higher immunity. The genomic mutation patterns of the high-risk and low-risk groups were significantly different, especially the KRAS mutation frequency was significantly higher in the high-risk group. Drug sensitivity analysis showed that the low-risk group was more sensitive to Erlotinib, Savolitinib and VE _ 822, which may be used as a potential drug for COAD treatment. Finally, the nomogram constructed by pathological features combined with RiskScore can accurately evaluate the prognosis of COAD patients. This study constructed and verified a TLS model that can predict COAD. More importantly, it provides a reference standard for guiding the prognosis and immunotherapy of COAD patients.

EMCMDA: predicting miRNA-disease associations via efficient matrix completion

Abundant researches have consistently illustrated the crucial role of microRNAs (miRNAs) in a wide array of essential biological processes. Furthermore, miRNAs have been validated as promising therapeutic targets for addressing complex diseases. Given the costly and time-consuming nature of traditional biological experimental validation methods, it is imperative to develop computational methods. In the work, we developed a novel approach named efficient matrix completion (EMCMDA) for predicting miRNA-disease associations. First, we calculated the similarities across multiple sources for miRNA/disease pairs and combined this information to create a holistic miRNA/disease similarity measure. Second, we utilized this biological information to create a heterogeneous network and established a target matrix derived from this network. Lastly, we framed the miRNA-disease association prediction issue as a low-rank matrix-complete issue that was addressed via minimizing matrix truncated schatten p-norm. Notably, we improved the conventional singular value contraction algorithm through using a weighted singular value contraction technique. This technique dynamically adjusts the degree of contraction based on the significance of each singular value, ensuring that the physical meaning of these singular values is fully considered. We evaluated the performance of EMCMDA by applying two distinct cross-validation experiments on two diverse databases, and the outcomes were statistically significant. In addition, we executed comprehensive case studies on two prevalent human diseases, namely lung cancer and breast cancer. Following prediction and multiple validations, it was evident that EMCMDA proficiently forecasts previously undisclosed disease-related miRNAs. These results underscore the robustness and efficacy of EMCMDA in miRNA-disease association prediction.

Expression analysis of necroptosis related genes and lncRNAs in patients with pituitary neuroendocrine tumors

Necroptosis can either be the cause of tumorigenesis or it can impede its process. Recently, it has been proved that long non-coding RNAs (lncRNAs) have different crucial roles at cellular level, especially on cell death. Regarding the important role of necroptosis and lncRNAs in the pathogenesis of different cancers, especially pituitary adenomas (PAs), we assessed expression levels of two necroptosis related genes, namely TRADD and BIRC2, in addition to three related lncRNAs, namely FLVCR1-DT, MAGI2-AS3, and NEAT1 in PAs compared with adjacent normal tissues (ANTs). TRADD had no significant difference between two groups; however, BIRC2, FLVCR1-DT, MAGI2-AS3, and NEAT1 were upregulated in PAs compared to ANTs (Expression ratios [95% CI] = 2.3 [1.47-3.6], 2.13 [1.02-4.44], 3.01 [1.76-5.16] and 2.47 [1.37-4.45], respectively). When taking into account different types of PAs, significant upregulation of BIRC2, MAGI2-AS3 and NEAT1 was recorded in non-functioning PAs compared with corresponding ANTs (Expression ratios [95% CI] =1.9 [1.04-3.43], 2.69 [1.26-5.72] and 2.22 [0.98-5.01], respectively). Additionally, higher levels of BIRC2 were associated with higher flow of CSF (P value=0.048). Moreover, higher Knosp classified tumors had lower levels of BIRC2 (P value=0.001). Finally, lower levels of MAGI2-AS3 were associated with larger tumor size (P value=0.006). NEAT1 expression was correlated with FLVCR1-DT and TRADD. TRADD expression was correlated with FLVCR1-DT. Additional correlation was observed between expression of BIRC2 and MAGI2-AS3. In sum, this study provides evidence that dysregulated levels of studied genes could contribute to the pathogenesis of pituitary tumors.

GEnDDn: An lncRNA-Disease Association Identification Framework Based on Dual-Net Neural Architecture and Deep Neural Network

Accumulating studies have demonstrated close relationships between long non-coding RNAs (lncRNAs) and diseases. Identification of new lncRNA-disease associations (LDAs) enables us to better understand disease mechanisms and further provides promising insights into cancer targeted therapy and anti-cancer drug design. Here, we present an LDA prediction framework called GEnDDn based on deep learning. GEnDDn mainly comprises two steps: First, features of both lncRNAs and diseases are extracted by combining similarity computation, non-negative matrix factorization, and graph attention auto-encoder, respectively. And each lncRNA-disease pair (LDP) is depicted as a vector based on concatenation operation on the extracted features. Subsequently, unknown LDPs are classified by aggregating dual-net neural architecture and deep neural network. Using six different evaluation metrics, we found that GEnDDn surpassed four competing LDA identification methods (SDLDA, LDNFSGB, IPCARF, LDASR) on the lncRNADisease and MNDR databases under fivefold cross-validation experiments on lncRNAs, diseases, LDPs, and independent lncRNAs and independent diseases, respectively. Ablation experiments further validated the powerful LDA prediction performance of GEnDDn. Furthermore, we utilized GEnDDn to find underlying lncRNAs for lung cancer and breast cancer. The results elucidated that there may be dense linkages between IFNG-AS1 and lung cancer as well as between HIF1A-AS1 and breast cancer. The results require further biomedical experimental verification. GEnDDn is publicly available at https://github.com/plhhnu/GEnDDn.

Computational prediction of new therapeutic effects of probiotics

Probiotics are living microorganisms that provide health benefits to their hosts, potentially aiding in the treatment or prevention of various diseases, including diarrhea, irritable bowel syndrome, ulcerative colitis, and Crohn's disease. Motivated by successful applications of link prediction in medical and biological networks, we applied link prediction to the probiotic-disease network to identify unreported relations. Using data from the Probio database and International Classification of Diseases-10th Revision (ICD-10) resources, we constructed a bipartite graph focused on the relationship between probiotics and diseases. We applied customized link prediction algorithms for this bipartite network, including common neighbors, Jaccard coefficient, and Adamic/Adar ranking formulas. We evaluated the results using Area under the Curve (AUC) and precision metrics. Our analysis revealed that common neighbors outperformed the other methods, with an AUC of 0.96 and precision of 0.6, indicating that basic formulas can predict at least six out of ten probable relations correctly. To support our findings, we conducted an exact search of the top 20 predictions and found six confirming papers on Google Scholar and Science Direct. Evidence suggests that Lactobacillus jensenii may provide prophylactic and therapeutic benefits for gastrointestinal diseases and that Lactobacillus acidophilus may have potential activity against urologic and female genital illnesses. Further investigation of other predictions through additional preclinical and clinical studies is recommended. Future research may focus on deploying more powerful link prediction algorithms to achieve better and more accurate results.

Improving plant miRNA-target prediction with self-supervised k-mer embedding and spectral graph convolutional neural network

Deciphering the targets of microRNAs (miRNAs) in plants is crucial for comprehending their function and the variation in phenotype that they cause. As the highly cell-specific nature of miRNA regulation, recent computational approaches usually utilize expression data to identify the most physiologically relevant targets. Although these methods are effective, they typically require a large sample size and high-depth sequencing to detect potential miRNA-target pairs, thereby limiting their applicability in improving plant breeding. In this study, we propose a novel miRNA-target prediction framework named kmerPMTF (k-mer-based prediction framework for plant miRNA-target). Our framework effectively extracts the latent semantic embeddings of sequences by utilizing k-mer splitting and a deep self-supervised neural network. We construct multiple similarity networks based on k-mer embeddings and employ graph convolutional networks to derive deep representations of miRNAs and targets and calculate the probabilities of potential associations. We evaluated the performance of kmerPMTF on four typical plant datasets: Arabidopsis thaliana, Oryza sativa, Solanum lycopersicum, and Prunus persica. The results demonstrate its ability to achieve AUPRC values of 84.9%, 91.0%, 80.1%, and 82.1% in 5-fold cross-validation, respectively. Compared with several state-of-the-art existing methods, our framework achieves better performance on threshold-independent evaluation metrics. Overall, our study provides an efficient and simplified methodology for identifying plant miRNA-target associations, which will contribute to a deeper comprehension of miRNA regulatory mechanisms in plants.

Identification of necroptosis-related gene signatures for predicting the prognosis of ovarian cancer

Ovarian cancer (OC) is one of the most prevalent and fatal malignant tumors of the female reproductive system. Our research aimed to develop a prognostic model to assist inclinical treatment decision-making.Utilizing data from The Cancer Genome Atlas (TCGA) and copy number variation (CNV) data from the University of California Santa Cruz (UCSC) database, we conducted analyses of differentially expressed genes (DEGs), gene function, and tumor microenvironment (TME) scores in various clusters of OC samples.Next, we classified participants into low-risk and high-risk groups based on the median risk score, thereby dividing both the training group and the entire group accordingly. Overall survival (OS) was significantly reduced in the high-risk group, and two independent prognostic factors were identified: age and risk score. Additionally, three genes-C-X-C Motif Chemokine Ligand 10 (CXCL10), RELB, and Caspase-3 (CASP3)-emerged as potential candidates for an independent prognostic signature with acceptable prognostic value. In Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses, pathways related to immune responses and inflammatory cell chemotaxis were identified. Cellular experiments further validated the reliability and precision of our findings. In conclusion, necroptosis-related genes play critical roles in tumor immunity, and our model introduces a novel strategy for predicting the prognosis of OC patients.

FM-FCN: A Neural Network with Filtering Modules for Accurate Vital Signs Extraction

Neural networks excel at capturing local spatial patterns through convolutional modules, but they may struggle to identify and effectively utilize the morphological and amplitude periodic nature of physiological signals. In this work, we propose a novel network named filtering module fully convolutional network (FM-FCN), which fuses traditional filtering techniques with neural networks to amplify physiological signals and suppress noise. First, instead of using a fully connected layer, we use an FCN to preserve the time-dimensional correlation information of physiological signals, enabling multiple cycles of signals in the network and providing a basis for signal processing. Second, we introduce the FM as a network module that adapts to eliminate unwanted interference, leveraging the structure of the filter. This approach builds a bridge between deep learning and signal processing methodologies. Finally, we evaluate the performance of FM-FCN using remote photoplethysmography. Experimental results demonstrate that FM-FCN outperforms the second-ranked method in terms of both blood volume pulse (BVP) signal and heart rate (HR) accuracy. It substantially improves the quality of BVP waveform reconstruction, with a decrease of 20.23% in mean absolute error (MAE) and an increase of 79.95% in signal-to-noise ratio (SNR). Regarding HR estimation accuracy, FM-FCN achieves a decrease of 35.85% in MAE, 29.65% in error standard deviation, and 32.88% decrease in 95% limits of agreement width, meeting clinical standards for HR accuracy requirements. The results highlight its potential in improving the accuracy and reliability of vital sign measurement through high-quality BVP signal extraction. The codes and datasets are available online at https://github.com/zhaoqi106/FM-FCN.

PBAC: A pathway-based attention convolution neural network for predicting clinical drug treatment responses

Precise and personalized drug application is crucial in the clinical treatment of complex diseases. Although neural networks offer a new approach to improving drug strategies, their internal structure is difficult to interpret. Here, we propose PBAC (Pathway-Based Attention Convolution neural network), which integrates a deep learning framework and attention mechanism to address the complex biological pathway information, thereby provide a biology function-based robust drug responsiveness prediction model. PBAC has four layers: gene-pathway layer, attention layer, convolution layer and fully connected layer. PBAC improves the performance of predicting drug responsiveness by focusing on important pathways, helping us understand the mechanism of drug action in diseases. We validated the PBAC model using data from four chemotherapy drugs (Bortezomib, Cisplatin, Docetaxel and Paclitaxel) and 11 immunotherapy datasets. In the majority of datasets, PBAC exhibits superior performance compared to traditional machine learning methods and other research approaches (area under curve = 0.81, the area under the precision-recall curve = 0.73). Using PBAC attention layer output, we identified some pathways as potential core cancer regulators, providing good interpretability for drug treatment prediction. In summary, we presented PBAC, a powerful tool to predict drug responsiveness based on the biology pathway information and explore the potential cancer-driving pathways.

Identification of biomarkers for abdominal aortic aneurysm in Behçet's disease via mendelian randomization and integrated bioinformatics analyses

Behçet's disease (BD) is a complex autoimmune disorder impacting several organ systems. Although the involvement of abdominal aortic aneurysm (AAA) in BD is rare, it can be associated with severe consequences. In the present study, we identified diagnostic biomarkers in patients with BD having AAA. Mendelian randomization (MR) analysis was initially used to explore the potential causal association between BD and AAA. The Limma package, WGCNA, PPI and machine learning algorithms were employed to identify potential diagnostic genes. A receiver operating characteristic curve (ROC) for the nomogram was constructed to ascertain the diagnostic value of AAA in patients with BD. Finally, immune cell infiltration analyses and single-sample gene set enrichment analysis (ssGSEA) were conducted. The MR analysis indicated a suggestive association between BD and the risk of AAA (odds ratio [OR]: 1.0384, 95% confidence interval [CI]: 1.0081-1.0696, p = 0.0126). Three hub genes (CD247, CD2 and CCR7) were identified using the integrated bioinformatics analyses, which were subsequently utilised to construct a nomogram (area under the curve [AUC]: 0.982, 95% CI: 0.944-1.000). Finally, the immune cell infiltration assay revealed that dysregulation immune cells were positively correlated with the three hub genes. Our MR analyses revealed a higher susceptibility of patients with BD to AAA. We used a systematic approach to identify three potential hub genes (CD247, CD2 and CCR7) and developed a nomogram to assist in the diagnosis of AAA among patients with BD. In addition, immune cell infiltration analysis indicated the dysregulation in immune cell proportions.

DP-GAN+B: A lightweight generative adversarial network based on depthwise separable convolutions for generating CT volumes

X-rays, commonly used in clinical settings, offer advantages such as low radiation and cost-efficiency. However, their limitation lies in the inability to distinctly visualize overlapping organs. In contrast, Computed Tomography (CT) scans provide a three-dimensional view, overcoming this drawback but at the expense of higher radiation doses and increased costs. Hence, from both the patient's and hospital's standpoints, there is substantial medical and practical value in attempting the reconstruction from two-dimensional X-ray images to three-dimensional CT images. In this paper, we introduce DP-GAN+B as a pioneering approach for transforming two-dimensional frontal and lateral lung X-rays into three-dimensional lung CT volumes. Our method innovatively employs depthwise separable convolutions instead of traditional convolutions and introduces vector and fusion loss for superior performance. Compared to prior models, DP-GAN+B significantly reduces the generator network parameters by 21.104 M and the discriminator network parameters by 10.82 M, resulting in a total reduction of 31.924 M (44.17%). Experimental results demonstrate that our network can effectively generate clinically relevant, high-quality CT images from X-ray data, presenting a promising solution for enhancing diagnostic imaging while mitigating cost and radiation concerns.

SEnSCA: Identifying possible ligand-receptor interactions and its application in cell-cell communication inference

Multicellular organisms have dense affinity with the coordination of cellular activities, which severely depend on communication across diverse cell types. Cell-cell communication (CCC) is often mediated via ligand-receptor interactions (LRIs). Existing CCC inference methods are limited to known LRIs. To address this problem, we developed a comprehensive CCC analysis tool SEnSCA by integrating single cell RNA sequencing and proteome data. SEnSCA mainly contains potential LRI acquisition and CCC strength evaluation. For acquiring potential LRIs, it first extracts LRI features and reduces the feature dimension, subsequently constructs negative LRI samples through K-means clustering, finally acquires potential LRIs based on Stacking ensemble comprising support vector machine, 1D-convolutional neural networks and multi-head attention mechanism. During CCC strength evaluation, SEnSCA conducts LRI filtering and then infers CCC by combining the three-point estimation approach and single cell RNA sequencing data. SEnSCA computed better precision, recall, accuracy, F1 score, AUC and AUPR under most of conditions when predicting possible LRIs. To better illustrate the inferred CCC network, SEnSCA provided three visualization options: heatmap, bubble diagram and network diagram. Its application on human melanoma tissue demonstrated its reliability in CCC detection. In summary, SEnSCA offers a useful CCC inference tool and is freely available at https://github.com/plhhnu/SEnSCA.

SCPLPA: An miRNA-disease association prediction model based on spatial consistency projection and label propagation algorithm

Identifying the association between miRNA and diseases is helpful for disease prevention, diagnosis and treatment. It is of great significance to use computational methods to predict potential human miRNA disease associations. Considering the shortcomings of existing computational methods, such as low prediction accuracy and weak generalization, we propose a new method called SCPLPA to predict miRNA-disease associations. First, a heterogeneous disease similarity network was constructed using the disease semantic similarity network and the disease Gaussian interaction spectrum kernel similarity network, while a heterogeneous miRNA similarity network was constructed using the miRNA functional similarity network and the miRNA Gaussian interaction spectrum kernel similarity network. Then, the estimated miRNA-disease association scores were evaluated by integrating the outcomes obtained by implementing label propagation algorithms in the heterogeneous disease similarity network and the heterogeneous miRNA similarity network. Finally, the spatial consistency projection algorithm of the network was used to extract miRNA disease association features to predict unverified associations between miRNA and diseases. SCPLPA was compared with four classical methods (MDHGI, NSEMDA, RFMDA and SNMFMDA), and the results of multiple evaluation metrics showed that SCPLPA exhibited the most outstanding predictive performance. Case studies have shown that SCPLPA can effectively identify miRNAs associated with colon neoplasms and kidney neoplasms. In summary, our proposed SCPLPA algorithm is easy to implement and can effectively predict miRNA disease associations, making it a reliable auxiliary tool for biomedical research.

Machine Learning Strategies in MicroRNA Research: Bridging Genome to Phenome

MicroRNAs (miRNAs) have emerged as a prominent layer of regulation of gene expression. This article offers the salient and current aspects of machine learning (ML) tools and approaches from genome to phenome in miRNA research. First, we underline that the complexity in the analysis of miRNA function ranges from their modes of biogenesis to the target diversity in diverse biological conditions. Therefore, it is imperative to first ascertain the miRNA coding potential of genomes and understand the regulatory mechanisms of their expression. This knowledge enables the efficient classification of miRNA precursors and the identification of their mature forms and respective target genes. Second, and because one miRNA can target multiple mRNAs and vice versa, another challenge is the assessment of the miRNA-mRNA target interaction network. Furthermore, long-noncoding RNA (lncRNA)and circular RNAs (circRNAs) also contribute to this complexity. ML has been used to tackle these challenges at the high-dimensional data level. The present expert review covers more than 100 tools adopting various ML approaches pertaining to, for example, (1) miRNA promoter prediction, (2) precursor classification, (3) mature miRNA prediction, (4) miRNA target prediction, (5) miRNA- lncRNA and miRNA-circRNA interactions, (6) miRNA-mRNA expression profiling, (7) miRNA regulatory module detection, (8) miRNA-disease association, and (9) miRNA essentiality prediction. Taken together, we unpack, critically examine, and highlight the cutting-edge synergy of ML approaches and miRNA research so as to develop a dynamic and microlevel understanding of human health and diseases.

Association between inflammation-based prognostic markers and mortality after hip replacement

We aimed to evaluate the association between inflammation-based prognostic markers and mortality after hip replacement. From March 2010 to June 2020, we identified 5,369 consecutive adult patients undergoing hip replacement with C-reactive protein (CRP), albumin, and complete blood count measured within six months before surgery. Receiver operating characteristic (ROC) curves were generated to evaluate predictabilities and estimate thresholds of CRP-to-albumin ratio (CAR), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR). Patients were divided according to threshold, and mortality risk was compared. The primary outcome was one-year mortality, and overall mortality was also analyzed. One-year mortality was 2.9%. Receiver operating characteristics analysis revealed areas under the curve of 0.838, 0.832, 0.701, and 0.732 for CAR, NLR, PLR, and modified Glasgow Prognostic Score, respectively. The estimated thresholds were 2.10, 3.16, and 11.77 for CAR, NLR, and PLR, respectively. According to the estimated threshold, high CAR and NLR were associated with higher one-year mortality after adjustment (1.0% vs. 11.7%; HR = 2.16; 95% CI 1.32-3.52; p = 0.002 for CAR and 0.8% vs. 9.6%; HR = 2.05; 95% CI 1.24-3.39; p = 0.01 for NLR), but PLR did not show a significant mortality increase (1.4% vs. 7.4%; HR = 1.12; 95% CI 0.77-1.63; p = 0.57). Our study demonstrated associations of preoperative levels of CAR and NLR with postoperative mortality in patients undergoing hip replacement. Our findings may be helpful in predicting mortality in patients undergoing hip replacement.

Identification of cuproptosis-realated key genes and pathways in Parkinson's disease via bioinformatics analysis

INTRODUCTION

Parkinson's disease (PD) is the second most common worldwide age-related neurodegenerative disorder without effective treatments. Cuproptosis is a newly proposed conception of cell death extensively studied in oncological diseases. Currently, whether cuproptosis contributes to PD remains largely unclear.

METHODS

The dataset GSE22491 was studied as the training dataset, and GSE100054 was the validation dataset. According to the expression levels of cuproptosis-related genes (CRGs) and differentially expressed genes (DEGs) between PD patients and normal samples, we obtained the differentially expressed CRGs. The protein-protein interaction (PPI) network was achieved through the Search Tool for the Retrieval of Interacting Genes. Meanwhile, the disease-associated module genes were screened from the weighted gene co-expression network analysis (WGCNA). Afterward, the intersection genes of WGCNA and PPI were obtained and enriched using the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). Subsequently, the key genes were identified from the datasets. The receiver operating characteristic curves were plotted and a PPI network was constructed, and the PD-related miRNAs and key genes-related miRNAs were intersected and enriched. Finally, the 2 hub genes were verified via qRT-PCR in the cell model of the PD and the control group.

RESULTS

525 DEGs in the dataset GSE22491 were identified, including 128 upregulated genes and 397 downregulated genes. Based on the PPI network, 41 genes were obtained. Additionally, the dataset was integrated into 34 modules by WGCNA. 36 intersection genes found from WGCNA and PPI were significantly abundant in 7 pathways. The expression levels of the genes were validated, and 2 key genes were obtained, namely peptidase inhibitor 3 (PI3) and neuroserpin family I member 1 (SERPINI1). PD-related miRNAs and key genes-related miRNAs were intersected into 29 miRNAs including hsa-miR-30c-2-3p. At last, the qRT-PCR results of 2 hub genes showed that the expressions of mRNA were up-regulated in PD.

CONCLUSION

Taken together, this study demonstrates the coordination of cuproptosis in PD. The key genes and miRNAs offer novel perspectives in the pathogenesis and molecular targeting treatment for PD.

A vector projection similarity-based method for miRNA-disease association prediction

[S U M M A R Y] Many miRNA-disease association prediction models incorporate Gaussian interaction profile kernel similarity (GIPS). However, the GIPS fails to consider the specificity of the miRNA-disease association matrix, where matrix elements with a value of 0 represent miRNA and disease relationships that have not been discovered yet. To address this issue and better account for the impact of known and unknown miRNA-disease associations on similarity, we propose a method called vector projection similarity-based method for miRNA-disease association prediction (VPSMDA). In VPSMDA, we introduce three projection rules and combined with logistic functions for the miRNA-disease association matrix and propose a vector projection similarity measure for miRNAs and diseases. By integrating the vector projection similarity matrix with the original one, we obtain the improved miRNA and disease similarity matrix. Additionally, we construct a weight matrix using different numbers of neighbors to reduce the noise in the similarity matrix. In performance evaluation, both LOOCV and 5-fold CV experiments demonstrate that VPSMDA outperforms seven other state-of-the-art methods in AUC. Furthermore, in a case study, VPSMDA successfully predicted 10, 9, and 10 out of the top 10 associations for three important human diseases, respectively, and these predictions were confirmed by recent biomedical resources.

Development and validation of prognostic nomographs for patients with cervical cancer: SEER-based Asian population study

To develop and validate a nomograph to predict the long-term survival probability of cervical cancer (CC) patients in Asia, Surveillance, Epidemiology, and End Results (SEER) were used to collect information about CC patients in Asia. The patient data were randomly sampled and divided into a training group and a validation group by 7:3. Least absolute shrinkage and selection operator (LASSO) regression was used to screen key indicators, and multivariate Cox regression model was used to establish a prognostic risk prediction model for CC patients. The receiver operating characteristic (ROC) curve and decision curve analysis (DCA) were adopted to comprehensively evaluate the nomogram model. LASSO regression and multivariate Cox proportional hazards model analysis showed that age, American Joint Committee on Cancer (AJCC) Stage, AJCC T, tumor size, and surgery were independent risk factors for prognosis. The ROC curve results proved that the area under curve (AUC) values of the training group in 3 and 5 years were 0.837 and 0.818, The AUC values of the validation group in 3 and 5 years were 0.796 and 0.783. DCA showed that the 3- and 5-year overall survival (OS) nomograms had good clinical potential value. The nomogram model developed in this study can effectively predict the prognosis of Asian patients with CC, and the risk stratification system based on this nomogram prediction model has some clinical value for discriminating high-risk patients.

Identification of circRNA-disease associations via multi-model fusion and ensemble learning

Circular RNA (circRNA) is a common non-coding RNA and plays an important role in the diagnosis and therapy of human diseases, circRNA-disease associations prediction based on computational methods can provide a new way for better clinical diagnosis. In this article, we proposed a novel method for circRNA-disease associations prediction based on ensemble learning, named ELCDA. First, the association heterogeneous network was constructed via collecting multiple information of circRNAs and diseases, and multiple similarity measures are adopted here, then, we use metapath, matrix factorization and GraphSAGE-based models to extract features of nodes from different views, the final comprehensive features of circRNAs and diseases via ensemble learning, finally, a soft voting ensemble strategy is used to integrate the predicted results of all classifier. The performance of ELCDA is evaluated by fivefold cross-validation and compare with other state-of-the-art methods, the experimental results show that ELCDA is outperformance than others. Furthermore, three common diseases are used as case studies, which also demonstrate that ELCDA is an effective method for predicting circRNA-disease associations.

MGCNSS: miRNA-disease association prediction with multi-layer graph convolution and distance-based negative sample selection strategy

Identifying disease-associated microRNAs (miRNAs) could help understand the deep mechanism of diseases, which promotes the development of new medicine. Recently, network-based approaches have been widely proposed for inferring the potential associations between miRNAs and diseases. However, these approaches ignore the importance of different relations in meta-paths when learning the embeddings of miRNAs and diseases. Besides, they pay little attention to screening out reliable negative samples which is crucial for improving the prediction accuracy. In this study, we propose a novel approach named MGCNSS with the multi-layer graph convolution and high-quality negative sample selection strategy. Specifically, MGCNSS first constructs a comprehensive heterogeneous network by integrating miRNA and disease similarity networks coupled with their known association relationships. Then, we employ the multi-layer graph convolution to automatically capture the meta-path relations with different lengths in the heterogeneous network and learn the discriminative representations of miRNAs and diseases. After that, MGCNSS establishes a highly reliable negative sample set from the unlabeled sample set with the negative distance-based sample selection strategy. Finally, we train MGCNSS under an unsupervised learning manner and predict the potential associations between miRNAs and diseases. The experimental results fully demonstrate that MGCNSS outperforms all baseline methods on both balanced and imbalanced datasets. More importantly, we conduct case studies on colon neoplasms and esophageal neoplasms, further confirming the ability of MGCNSS to detect potential candidate miRNAs. The source code is publicly available on GitHub https://github.com/15136943622/MGCNSS/tree/master.

Prognostic mutation signature would serve as a potential prognostic predictor in patients with diffuse large B-cell lymphoma

The present study aimed to elucidate the prognostic mutation signature (PMS) associated with long-term survival in a diffuse large B-cell lymphoma (DLBCL) cohort. All data including derivation and validation cohorts were retrospectively retrieved from The Cancer Genome Atlas (TCGA) database and whole-exome sequencing (WES) data. The Lasso Cox regression analysis was used to construct the PMS based on WES data, and the PMS was determined using the area under the receiver operating curve (AUC). The predictive performance of eligible PMS was analyzed by time-dependent receiver operating curve (ROC) analyses. After the initial evaluation, a PMS composed of 94 PFS-related genes was constructed. Notably, this constructed PMS accurately predicted the 12-, 36-, and 60-month PFS, with AUC values of 0.982, 0.983, and 0.987, respectively. A higher level of PMS was closely linked to a significantly worse PFS, regardless of the molecular subtype. Further evaluation by forest plot revealed incorporation of international prognostic index or tumor mutational burden into PMS increased the prediction capability for PFS. The drug-gene interaction and pathway exploration revealed the PFS-related genes were associated with DNA damage, TP53, apoptosis, and immune cell functions. In conclusion, this study utilizing a high throughput genetic approach demonstrated that the PMS could serve as a prognostic predictor in DLBCL patients. Furthermore, the identification of the key signaling pathways for disease progression also provides information for further investigation to gain more insight into novel drug-resistant mechanisms.

Prognostic model revealing pyroptosis-related signatures in oral squamous cell carcinoma based on bioinformatics analysis

One of the most common oral carcinomas is oral squamous cell carcinoma (OSCC), bringing a heavy burden to global health. Although progresses have been made in the intervention of OSCC, 5 years survival of patients suffering from OSCC is poor like before regarding to the high invasiveness of OSCC, which causes metastasis and recurrence of the tumor. The relationship between pyroptosis and OSCC remains to be further investigated as pyroptosis in carcinomas has gained much attention. Herein, the key pyroptosis-related genes were identified according to The Cancer Genome Atlas (TCGA) dataset. Additionally, a prognostic model was constructed based upon three key genes (CTLA4, CD5, and IL12RB2) through least absolute shrinkage and selection operator (LASSO) analyses, as well as univariate and multivariate COX regression in OSCC. It was discovered that the high expression of these three genes was associated with the low-risk group. We also identified LAIR2 as a hub gene, whose expression negatively correlated with the risk score and the different immune cell infiltration. Finally, we proved that these three genes were independent prognostic factors linked to overall survival (OS), and reliable consequences could be predicted by this model. Our study revealed the relationship between pyroptosis and OSCC, providing insights into new treatment targets for preventing and treating OSCC.

A prognostic nomogram for recurrence survival in post-surgical patients with varicose veins of the lower extremities

Varicose veins of the lower extremities (VVLEs) are prevalent globally. This study aims to identify prognostic factors and develop a prediction model for recurrence survival (RS) in VVLEs patients after surgery. A retrospective analysis of VVLEs patients from the Third Hospital of Nanchang was conducted between April 2017 and March 2022. A LASSO (Least Absolute Shrinkage and Selection Operator) regression model pinpointed significant recurrence predictors, culminating in a prognostic nomogram. The model's performance was evaluated by C-index, receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). The LASSO regression identified seven predictors for the nomogram predicting 1-, 2-, and 5-year RS. These predictors were age, body mass index (BMI), hypertension, diabetes, the Clinical Etiological Anatomical Pathophysiological (CEAP) grade, iliac vein compression syndrome (IVCS), and postoperative compression stocking duration (PCSD). The nomogram's C-index was 0.716, with AUCs (Area Under the Curve scores) of 0.705, 0.725, and 0.758 for 1-, 2-, and 5-year RS, respectively. Calibration and decision curve analyses validated the model's predictive accuracy and clinical utility. Kaplan-Meier analysis distinguished between low and high-risk groups with significant prognostic differences (P < 0.05). This study has successfully developed and validated a nomogram for predicting RS in patients with VVLEs after surgery, enhancing personalized care and informing clinical decision-making.

Finding potential lncRNA-disease associations using a boosting-based ensemble learning model

Introduction: Long non-coding RNAs (lncRNAs) have been in the clinical use as potential prognostic biomarkers of various types of cancer. Identifying associations between lncRNAs and diseases helps capture the potential biomarkers and design efficient therapeutic options for diseases. Wet experiments for identifying these associations are costly and laborious. Methods: We developed LDA-SABC, a novel boosting-based framework for lncRNA-disease association (LDA) prediction. LDA-SABC extracts LDA features based on singular value decomposition (SVD) and classifies lncRNA-disease pairs (LDPs) by incorporating LightGBM and AdaBoost into the convolutional neural network. Results: The LDA-SABC performance was evaluated under five-fold cross validations (CVs) on lncRNAs, diseases, and LDPs. It obviously outperformed four other classical LDA inference methods (SDLDA, LDNFSGB, LDASR, and IPCAF) through precision, recall, accuracy, F1 score, AUC, and AUPR. Based on the accurate LDA prediction performance of LDA-SABC, we used it to find potential lncRNA biomarkers for lung cancer. The results elucidated that 7SK and HULC could have a relationship with non-small-cell lung cancer (NSCLC) and lung adenocarcinoma (LUAD), respectively. Conclusion: We hope that our proposed LDA-SABC method can help improve the LDA identification.

Identification of crosstalk genes relating to ECM-receptor interaction genes in MASH and DN using bioinformatics and machine learning

This study aimed to identify genes shared by metabolic dysfunction-associated fatty liver disease (MASH) and diabetic nephropathy (DN) and the effect of extracellular matrix (ECM) receptor interaction genes on them. Datasets with MASH and DN were downloaded from the Gene Expression Omnibus (GEO) database. Pearson's coefficients assessed the correlation between ECM-receptor interaction genes and cross talk genes. The coexpression network of co-expression pairs (CP) genes was integrated with its protein-protein interaction (PPI) network, and machine learning was employed to identify essential disease-representing genes. Finally, immuno-penetration analysis was performed on the MASH and DN gene datasets using the CIBERSORT algorithm to evaluate the plausibility of these genes in diseases. We found 19 key CP genes. Fos proto-oncogene (FOS), belonging to the IL-17 signalling pathway, showed greater centrality PPI network; Hyaluronan Mediated Motility Receptor (HMMR), belonging to ECM-receptor interaction genes, showed most critical in the co-expression network map of 19 CP genes; Forkhead Box C1 (FOXC1), like FOS, showed a high ability to predict disease in XGBoost analysis. Further immune infiltration showed a clear positive correlation between FOS/FOXC1 and mast cells that secrete IL-17 during inflammation. Combining the results of previous studies, we suggest a FOS/FOXC1/HMMR regulatory axis in MASH and DN may be associated with mast cells in the acting IL-17 signalling pathway. Extracellular HMMR may regulate the IL-17 pathway represented by FOS through the Mitogen-Activated Protein Kinase 1 (ERK) or PI3K-Akt-mTOR pathway. HMMR may serve as a signalling carrier between MASH and DN and could be targeted for therapeutic development.

Prediction of MicroRNA-Disease Potential Association Based on Sparse Learning and Multilayer Random Walks

More and more studies have shown that microRNAs (miRNAs) play an indispensable role in the study of complex diseases in humans. Traditional biological experiments to detect miRNA-disease associations are expensive and time-consuming. Therefore, it is necessary to propose efficient and meaningful computational models to predict miRNA-disease associations. In this study, we aim to propose a miRNA-disease association prediction model based on sparse learning and multilayer random walks (SLMRWMDA). The miRNA-disease association matrix is decomposed and reconstructed by the sparse learning method to obtain richer association information, and at the same time, the initial probability matrix for the random walk with restart algorithm is obtained. The disease similarity network, miRNA similarity network, and miRNA-disease association network are used to construct heterogeneous networks, and the stable probability is obtained based on the topological structure features of diseases and miRNAs through a multilayer random walk algorithm to predict miRNA-disease potential association. The experimental results show that the prediction accuracy of this model is significantly improved compared with the previous related models. We evaluated the model using global leave-one-out cross-validation (global LOOCV) and fivefold cross-validation (5-fold CV). The area under the curve (AUC) value for the LOOCV is 0.9368. The mean AUC value for 5-fold CV is 0.9335 and the variance is 0.0004. In the case study, the results show that SLMRWMDA is effective in inferring the potential association of miRNA-disease.

HormoNet: a deep learning approach for hormone-drug interaction prediction

Several experimental evidences have shown that the human endogenous hormones can interact with drugs in many ways and affect drug efficacy. The hormone drug interactions (HDI) are essential for drug treatment and precision medicine; therefore, it is essential to understand the hormone-drug associations. Here, we present HormoNet to predict the HDI pairs and their risk level by integrating features derived from hormone and drug target proteins. To the best of our knowledge, this is one of the first attempts to employ deep learning approach for prediction of HDI prediction. Amino acid composition and pseudo amino acid composition were applied to represent target information using 30 physicochemical and conformational properties of the proteins. To handle the imbalance problem in the data, we applied synthetic minority over-sampling technique technique. Additionally, we constructed novel datasets for HDI prediction and the risk level of their interaction. HormoNet achieved high performance on our constructed hormone-drug benchmark datasets. The results provide insights into the understanding of the relationship between hormone and a drug, and indicate the potential benefit of reducing risk levels of interactions in designing more effective therapies for patients in drug treatments. Our benchmark datasets and the source codes for HormoNet are available in: https://github.com/EmamiNeda/HormoNet .

Bioinformatics analysis for the identification of Sprouty-related EVH1 domain-containing protein 3 expression and its clinical significance in thyroid carcinoma

The poorly differentiated thyroid carcinoma (THCA) subtype is associated with an aggressive disease course, a less favorable overall prognosis, and an increased risk of distant organ metastasis. In this study, our objective was to explore the potential utility of the Sprouty-related EVH1 domain-containing protein 3 (SPRED3) as a biomarker for early diagnosis and prognosis in THCA patients. The differentially expressed prognostic-related genes associated with THCA were identified by querying The Cancer Genome Atlas (TCGA) database. The difference in the expression of the SPRED3 gene between thyroid carcinoma (THCA) tissues and normal tissues was analyzed using data from The Cancer Genome Atlas (TCGA) and further validated through immunohistochemistry. Univariate and multivariate Cox regression models were used, along with clinical information from THCA patients, to analyze the prognostic value of the SPRED3 gene in THCA patients. Functional enrichment analysis was subsequently performed to elucidate the molecular mechanisms underlying the regulatory effects of the SPRED3 gene on thyroid carcinoma. Additionally, we calculated the percentage of infiltrating immune cells in THCA patients and evaluated their correlation with SPRED3 gene expression. Compared with those in noncancerous thyroid tissue, the gene and protein expression levels of SPRED3 were found to be elevated in thyroid carcinoma tissues. Furthermore, the expression of SPRED3 in thyroid carcinoma exhibited significant correlations with tumor location, histological grade, pathological stage, and tumor node metastasis classification (TNM) stage. Univariate and multivariate Cox proportional hazards (Cox) regression analyses demonstrated that SPRED3 could serve as an independent prognostic factor for predicting the overall survival of THCA patients. The results of functional enrichment analysis suggested the potential involvement of SPRED3 in the regulation of extracellular matrix organization, epidermal development, signaling receptor activator activity, skin development, receptor ligand activity, glycosaminoglycan binding, neuroactive ligand‒receptor interaction, the IL-17 signaling pathway, and the PI3K-Akt signaling pathway. Additionally, there were significant correlations between the expression level of the SPRED3 gene and the infiltration of various immune cells (eosinophils, central memory T cells, neutrophils, macrophages, and NK cells) within the thyroid tumor microenvironment. SPRED3 can be used as a prognostic biomarker in patients with THCA could potentially be therapeutic target for THCA.

Leveraging a meta-learning approach to advance the accuracy of Nav blocking peptides prediction

The voltage-gated sodium (Nav) channel is a crucial molecular component responsible for initiating and propagating action potentials. While the α subunit, forming the channel pore, plays a central role in this function, the complete physiological function of Nav channels relies on crucial interactions between the α subunit and auxiliary proteins, known as protein-protein interactions (PPI). Nav blocking peptides (NaBPs) have been recognized as a promising and alternative therapeutic agent for pain and itch. Although traditional experimental methods can precisely determine the effect and activity of NaBPs, they remain time-consuming and costly. Hence, machine learning (ML)-based methods that are capable of accurately contributing in silico prediction of NaBPs are highly desirable. In this study, we develop an innovative meta-learning-based NaBP prediction method (MetaNaBP). MetaNaBP generates new feature representations by employing a wide range of sequence-based feature descriptors that cover multiple perspectives, in combination with powerful ML algorithms. Then, these feature representations were optimized to identify informative features using a two-step feature selection method. Finally, the selected informative features were applied to develop the final meta-predictor. To the best of our knowledge, MetaNaBP is the first meta-predictor for NaBP prediction. Experimental results demonstrated that MetaNaBP achieved an accuracy of 0.948 and a Matthews correlation coefficient of 0.898 over the independent test dataset, which were 5.79% and 11.76% higher than the existing method. In addition, the discriminative power of our feature representations surpassed that of conventional feature descriptors over both the training and independent test datasets. We anticipate that MetaNaBP will be exploited for the large-scale prediction and analysis of NaBPs to narrow down the potential NaBPs.

Identifying CTH and MAP1LC3B as ferroptosis biomarkers for prognostic indication in gastric cancer decoding

Gastric cancer (GC), known for its high incidence and poor prognosis, urgently necessitates the identification of reliable prognostic biomarkers to enhance patient outcomes. We scrutinized data from 375 GC patients alongside 32 non-cancer controls, sourced from the TCGA database. A univariate Cox Proportional Hazards Model (COX) regression was employed to evaluate expressions of ferroptosis-related genes. This was followed by the application of Least Absolute Shrinkage and Selection Operator (LASSO) and multivariate COX regression for the development of prognostic models. The composition of immune cell subtypes was quantified utilizing CIBERSORT, with their distribution in GC versus control samples being comparatively analyzed. Furthermore, the correlation between the expressions of Cystathionine Gamma-Lyase (CTH) and Microtubule Associated Protein 1 Light Chain 3 Beta (MAP1LC3B) and the abundance of immune cell subtypes was explored. Our bioinformatics findings underwent validation through immunohistochemical analysis. Our prognostic models integrated CTH and MAP1LC3B. Survival analysis indicated that patients categorized as high-risk, as defined by the model, exhibited significantly lower survival rates compared to their low-risk counterparts. Notably, CTH expression inversely correlated with monocyte levels, while MAP1LC3B expression showed an inverse relationship with the abundance of M2 macrophages. Immunohistochemical validation corroborated lower expressions of CTH and MAP1LC3B in GC tissues relative to control samples, in concordance with our bioinformatics predictions. Our study suggests that the dysregulation of CTH, MAP1LC3B, and the accompanying monocyte-macrophage dynamics could be pivotal in the prognosis of GC. These elements present potential targets for prognostic assessment and therapeutic intervention.

Exploring the therapeutic mechanisms and prognostic targets of Biochanin A in glioblastoma via integrated computational analysis and in vitro experiments

Glioblastoma (GBM) is the most aggressive brain tumor and is characterized by a poor prognosis and high recurrence and mortality rates. Biochanin A (BCA) exhibits promising clinical anti-tumor effects. In this study, we aimed to explore the pharmacological mechanisms by which BCA acts against GBM. Network pharmacology was employed to identify overlapping target genes between BCA and GBM. Differentially expressed genes from the Gene Expression Profiling Interactive Analysis 2 (GEPIA2) database were visualized using VolcaNose. Interactions among these overlapping genes were analyzed using the Search Tool for the Retrieval of Interacting Genes/Proteins database. Protein-protein interaction networks were constructed using Cytoscape 3.8.1. The Kyoto Encyclopedia of Genes and Genomes pathway and Gene Ontology enrichment analyses were conducted using the Database for Annotation, Visualization, and Integrated Discovery. Survival analyses for these genes were performed using the GEPIA2 database. The Chinese Glioma Genome Atlas database was used to study the correlations between key prognostic genes. Molecular docking was confirmed using the DockThor database and visualized with PyMol software. Cell viability was assessed via the CCK-8 assay, apoptosis and the cell cycle stages were examined using flow cytometry, and protein expression was detected using western blotting. In all, 63 genes were initially identified as potential targets for BCA in treating GBM. Enrichment analysis suggested that the pharmacological mechanisms of BCA primarily involved cell cycle inhibition, induction of cell apoptosis, and immune regulation. Based on these findings, AKT1, EGFR, CASP3, and MMP9 were preliminarily predicted as key prognostic target genes for BCA in GBM treatment. Furthermore, molecular docking analysis suggested stable binding of BCA to the target protein. In vitro experiments revealed the efficacy of BCA in inhibiting GBM, with an IC50 value of 98.37 ± 2.21 μM. BCA inhibited cell proliferation, induced cell apoptosis, and arrested the cell cycle of GBM cells. Furthermore, the anti-tumor effects of BCA on U251 cells were linked to the regulation of the target protein. We utilized integrated bioinformatics analyses to predict targets and confirmed through experiments that BCA possesses remarkable anti-tumor activities. We present a novel approach for multi-target treatment of GBM using BCA.

The metabolism-related lncRNA signature predicts the prognosis of breast cancer patients

Long non-coding RNAs (lncRNAs) involved in metabolism are recognized as significant factors in breast cancer (BC) progression. We constructed a novel prognostic signature for BC using metabolism-related lncRNAs and investigated their underlying mechanisms. The training and validation cohorts were established from BC patients acquired from two public sources: The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). The prognostic signature of metabolism-related lncRNAs was constructed using the least absolute shrinkage and selection operator (LASSO) cox regression analysis. We developed and validated a new prognostic risk model for BC using the signature of metabolism-related lncRNAs (SIRLNT, SIAH2-AS1, MIR205HG, USP30-AS1, MIR200CHG, TFAP2A-AS1, AP005131.2, AL031316.1, C6orf99). The risk score obtained from this signature was proven to be an independent prognostic factor for BC patients, resulting in a poor overall survival (OS) for individuals in the high-risk group. The area under the curve (AUC) for OS at three and five years were 0.67 and 0.65 in the TCGA cohort, and 0.697 and 0.68 in the GEO validation cohort, respectively. The prognostic signature demonstrated a robust association with the immunological state of BC patients. Conventional chemotherapeutics, such as docetaxel and paclitaxel, showed greater efficacy in BC patients classified as high-risk. A nomogram with a c-index of 0.764 was developed to forecast the survival time of BC patients, considering their risk score and age. The silencing of C6orf99 markedly decreased the proliferation, migration, and invasion capacities in MCF-7 cells. Our study identified a signature of metabolism-related lncRNAs that predicts outcomes in BC patients and could assist in tailoring personalized prevention and treatment plans.

CarSitePred: an integrated algorithm for identifying carbonylated sites based on KNDUA-LNDOT resampling technique

Carbonylated sites are the determining factors for functional changes or deletions in carbonylated proteins, so identifying carbonylated sites is essential for understanding the process of protein carbonylated and exploring the pathogenesis of related diseases. The current wet experimental methods for predicting carbonylated modification sites ae not only expensive and time-consuming, but also have limited protein processing capabilities and cannot meet the needs of researchers. The identification of carbonylated sites using computational methods not only improves the functional characterization of proteins, but also provides researchers with free tools for predicting carbonylated sites. Therefore, it is essential to establish a model using computational methods that can accurately predict protein carbonylated sites. In this study, a prediction model, CarSitePred, is proposed to identify carbonylation sites. In CarSitePred, specific location amino acid hydrophobic hydrophilic, one-to-one numerical conversion of amino acids, and AlexNet convolutional neural networks convert preprocessed carbonylated sequences into valid numerical features. The K-means Normal Distribution-based Undersampling Algorithm (KNDUA) and Localized Normal Distribution Oversampling Technology (LNDOT) were firstly proposed and employed to balance the K, P, R and T carbonylation training dataset. And for the first time, carbonylation modification sites were transformed into the form of images and directly inputted into AlexNet convolutional neural network to extract features for fitting SVM classifiers. The 10-fold cross-validation and independent testing results show that CarSitePred achieves better prediction performance than the best currently available prediction models. Availability: https://github.com/zuoyun123/CarSitePred.

Oral_voting_transfer: classification of oral microorganisms' function proteins with voting transfer model

Introduction

The oral microbial group typically represents the human body's highly complex microbial group ecosystem. Oral microorganisms take part in human diseases, including Oral cavity inflammation, mucosal disease, periodontal disease, tooth decay, and oral cancer. On the other hand, oral microbes can also cause endocrine disorders, digestive function, and nerve function disorders, such as diabetes, digestive system diseases, and Alzheimer's disease. It was noted that the proteins of oral microbes play significant roles in these serious diseases. Having a good knowledge of oral microbes can be helpful in analyzing the procession of related diseases. Moreover, the high-dimensional features and imbalanced data lead to the complexity of oral microbial issues, which can hardly be solved with traditional experimental methods.

Methods

To deal with these challenges, we proposed a novel method, which is oral_voting_transfer, to deal with such classification issues in the field of oral microorganisms. Such a method employed three features to classify the five oral microorganisms, including Streptococcus mutans, Staphylococcus aureus, abiotrophy adjacent, bifidobacterial, and Capnocytophaga. Firstly, we utilized the highly effective model, which successfully classifies the organelle's proteins and transfers to deal with the oral microorganisms. And then, some classification methods can be treated as the local classifiers in this work. Finally, the results are voting from the transfer classifiers and the voting ones.

Results and discussion

The proposed method achieved the well performances in the five oral microorganisms. The oral_voting_transfer is a standalone tool, and all its source codes are publicly available at https://github.com/baowz12345/voting_transfer.

Deep-WET: a deep learning-based approach for predicting DNA-binding proteins using word embedding techniques with weighted features

DNA-binding proteins (DBPs) play a significant role in all phases of genetic processes, including DNA recombination, repair, and modification. They are often utilized in drug discovery as fundamental elements of steroids, antibiotics, and anticancer drugs. Predicting them poses the most challenging task in proteomics research. Conventional experimental methods for DBP identification are costly and sometimes biased toward prediction. Therefore, developing powerful computational methods that can accurately and rapidly identify DBPs from sequence information is an urgent need. In this study, we propose a novel deep learning-based method called Deep-WET to accurately identify DBPs from primary sequence information. In Deep-WET, we employed three powerful feature encoding schemes containing Global Vectors, Word2Vec, and fastText to encode the protein sequence. Subsequently, these three features were sequentially combined and weighted using the weights obtained from the elements learned through the differential evolution (DE) algorithm. To enhance the predictive performance of Deep-WET, we applied the SHapley Additive exPlanations approach to remove irrelevant features. Finally, the optimal feature subset was input into convolutional neural networks to construct the Deep-WET predictor. Both cross-validation and independent tests indicated that Deep-WET achieved superior predictive performance compared to conventional machine learning classifiers. In addition, in extensive independent test, Deep-WET was effective and outperformed than several state-of-the-art methods for DBP prediction, with accuracy of 78.08%, MCC of 0.559, and AUC of 0.805. This superior performance shows that Deep-WET has a tremendous predictive capacity to predict DBPs. The web server of Deep-WET and curated datasets in this study are available at https://deepwet-dna.monarcatechnical.com/ . The proposed Deep-WET is anticipated to serve the community-wide effort for large-scale identification of potential DBPs.

Identification and analysis of key circRNAs in the mouse embryonic ovary provides insight into primordial follicle development

BACKGROUND

CircRNAs are a class of noncoding RNAs with tissue- and development-specific expression characteristics. In many mammals, primordial follicle development begins in the embryonic stage. However, the study of circRNAs in primordial follicle development in mice has not been reported.

RESULTS

In this study, ovaries were collected from mouse foetuses at 15.5 days post coitus (dpc) and 17.5 dpc, which are two key stages of primordial follicle development. A total of 4785 circRNAs were obtained by using RNA-seq. Of these, 83 differentially expressed circRNAs were identified. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses showed that these differential circRNAs were mainly involved in the regulation of reproductive development. Through qRT-PCR, back-splice sequence detection and enzyme digestion protection experiments, we found that circ-009346, circ-014674, circ-017054 and circ-008296 were indeed circular. Furthermore, circ-009346, circ-014674 and circ-017054 were identified as three key circRNAs by analysing their expression in the ovaries of mice at different developmental stages. The circRNA-miRNA-mRNA interaction network was constructed and validated for target miRNA and mRNA using qRT-PCR. The interacting genes circ-009346, circ-014674, and circ-017054 were subjected to KEGG enrichment analysis. We found that circ-014674 may participate in the assembly and reserve of primordial follicles through oestrogen and the Janus kinase (JAK) signal transducer and activator of transcription (STAT) signalling pathway (JAK-SATA). Circ-009346 and circ-017054 may have similar functions and are involved in the activation and growth of primordial follicles through the mitogen-activated protein kinase (MAPK) and phosphoinositide 3-kinase (PI3K) signalling pathways.

CONCLUSIONS

Based on our findings, three circRNAs associated with primordial follicle development were identified, and their potential mechanisms of regulating primordial follicle development were revealed. These findings will help us better understand the molecular mechanism of circRNAs in primordial follicles and provide important references and targets for the development of primordial follicles.

MLFLHMDA: predicting human microbe-disease association based on multi-view latent feature learning

Introduction

A growing body of research indicates that microorganisms play a crucial role in human health. Imbalances in microbial communities are closely linked to human diseases, and identifying potential relationships between microbes and diseases can help elucidate the pathogenesis of diseases. However, traditional methods based on biological or clinical experiments are costly, so the use of computational models to predict potential microbe-disease associations is of great importance.

Methods

In this paper, we present a novel computational model called MLFLHMDA, which is based on a Multi-View Latent Feature Learning approach to predict Human potential Microbe-Disease Associations. Specifically, we compute Gaussian interaction profile kernel similarity between diseases and microbes based on the known microbe-disease associations from the Human Microbe-Disease Association Database and perform a preprocessing step on the resulting microbe-disease association matrix, namely, weighting K nearest known neighbors (WKNKN) to reduce the sparsity of the microbe-disease association matrix. To obtain unobserved associations in the microbe and disease views, we extract different latent features based on the geometrical structure of microbes and diseases, and project multi-modal latent features into a common subspace. Next, we introduce graph regularization to preserve the local manifold structure of Gaussian interaction profile kernel similarity and add L p , q -norms to the projection matrix to ensure the interpretability and sparsity of the model.

Results

The AUC values for global leave-one-out cross-validation and 5-fold cross validation implemented by MLFLHMDA are 0.9165 and 0.8942+/-0.0041, respectively, which perform better than other existing methods. In addition, case studies of different diseases have demonstrated the superiority of the predictive power of MLFLHMDA. The source code of our model and the data are available on https://github.com/LiangzheZhang/MLFLHMDA_master.

Multi-scale topology and position feature learning and relationship-aware graph reasoning for prediction of drug-related microbes

MOTIVATION

The human microbiome may impact the effectiveness of drugs by modulating their activities and toxicities. Predicting candidate microbes for drugs can facilitate the exploration of the therapeutic effects of drugs. Most recent methods concentrate on constructing of the prediction models based on graph reasoning. They fail to sufficiently exploit the topology and position information, the heterogeneity of multiple types of nodes and connections, and the long-distance correlations among nodes in microbe-drug heterogeneous graph.

RESULTS

We propose a new microbe-drug association prediction model, NGMDA, to encode the position and topological features of microbe (drug) nodes, and fuse the different types of features from neighbors and the whole heterogeneous graph. First, we formulate the position and topology features of microbe (drug) nodes by t-step random walks, and the features reveal the topological neighborhoods at multiple scales and the position of each node. Second, as the features of nodes are high-dimensional and sparse, we designed an embedding enhancement strategy based on supervised fully connected autoencoders to form the embeddings with representative features and the more discriminative node distributions. Third, we propose an adaptive neighbor feature fusion module, which fuses features of neighbors by the constructed position- and topology-sensitive heterogeneous graph neural networks. A novel self-attention mechanism is developed to estimate the importance of the position and topology of each neighbor to a target node. Finally, a heterogeneous graph feature fusion module is constructed to learn the long-distance correlations among the nodes in the whole heterogeneous graph by a relationship-aware graph transformer. Relationship-aware graph transformer contains the strategy for encoding the connection relationship types among the nodes, which is helpful for integrating the diverse semantics of these connections. The extensive comparison experimental results demonstrate NGMDA's superior performance over five state-of-the-art prediction methods. The ablation experiment shows the contributions of the multi-scale topology and position feature learning, the embedding enhancement strategy, the neighbor feature fusion, and the heterogeneous graph feature fusion. Case studies over three drugs further indicate that NGMDA has ability in discovering the potential drug-related microbes.

AVAILABILITY AND IMPLEMENTATION

Source codes and Supplementary Material are available at https://github.com/pingxuan-hlju/NGMDA.

Prediction of miRNA-disease associations based on strengthened hypergraph convolutional autoencoder

Most existing graph neural network-based methods for predicting miRNA-disease associations rely on initial association matrices to pass messages, but the sparsity of these matrices greatly limits performance. To address this issue and predict potential associations between miRNAs and diseases, we propose a method called strengthened hypergraph convolutional autoencoder (SHGAE). SHGAE leverages multiple layers of strengthened hypergraph neural networks (SHGNN) to obtain robust node embeddings. Within SHGNN, we design a strengthened hypergraph convolutional network module (SHGCN) that enhances original graph associations and reduces matrix sparsity. Additionally, SHGCN expands node receptive fields by utilizing hyperedge features as intermediaries to obtain high-order neighbor embeddings. To improve performance, we also incorporate attention-based fusion of self-embeddings and SHGCN embeddings. SHGAE predicts potential miRNA-disease associations using a multilayer perceptron as the decoder. Across multiple metrics, SHGAE outperforms other state-of-the-art methods in five-fold cross-validation. Furthermore, we evaluate SHGAE on colon and lung neoplasms cases to demonstrate its ability to predict potential associations. Notably, SHGAE also performs well in the analysis of gastric neoplasms without miRNA associations.

Identification of the shared genes in type 2 diabetes mellitus and osteoarthritis and the role of quercetin

This study investigated the underlying comorbidity mechanism between type 2 diabetes mellitus (T2DM) and osteoarthritis (OA), while also assessing the therapeutic potential of quercetin for early intervention and treatment of these two diseases. The shared genes were obtained through GEO2R, limma and weighted gene co-expression network analysis (WGCNA), and validated using clinical databases and the area under the curves (ROC). Functional enrichment analysis was conducted to elucidate the underlying mechanisms of comorbidity between T2DM and OA. The infiltration of immune cells was analysed using the CIBERSORT algorithm in conjunction with ESTIMATE algorithm. Subsequently, transcriptional regulation analysis, potential chemical prediction, gene-disease association, relationships between the shared genes and ferroptosis as well as immunity-related genes were investigated along with molecular docking. We identified the 12 shared genes (EPHA3, RASIP1, PENK, LRRC17, CEBPB, EFEMP2, UBAP1, PPP1R15A, SPEN, MAFF, GADD45B and KLF4) across the four datasets. Our predictions suggested that targeting these shared genes could potentially serve as therapeutic interventions for both T2DM and OA. Specifically, they are involved in key signalling pathways such as p53, IL-17, NF-kB and MAPK signalling pathways. Furthermore, the regulation of ferroptosis and immunity appears to be interconnected in both diseases. Notably, in this context quercetin emerges as a promising drug candidate for treating T2DM and OA by specifically targeting the shared genes. We conducted a bioinformatics analysis to identify potential therapeutic targets, mechanisms and drugs for T2DM and OA, thereby offering novel insights into molecular therapy for these two diseases.

A novel diabetic foot ulcer diagnostic model: identification and analysis of genes related to glutamine metabolism and immune infiltration

BACKGROUND

Diabetic foot ulcer (DFU) is one of the most common and severe complications of diabetes, with vascular changes, neuropathy, and infections being the primary pathological mechanisms. Glutamine (Gln) metabolism has been found to play a crucial role in diabetes complications. This study aims to identify and validate potential Gln metabolism biomarkers associated with DFU through bioinformatics and machine learning analysis.

METHODS

We downloaded two microarray datasets related to DFU patients from the Gene Expression Omnibus (GEO) database, namely GSE134431, GSE68183, and GSE80178. From the GSE134431 dataset, we obtained differentially expressed Gln-metabolism related genes (deGlnMRGs) between DFU and normal controls. We analyzed the correlation between deGlnMRGs and immune cell infiltration status. We also explored the relationship between GlnMRGs molecular clusters and immune cell infiltration status. Notably, WGCNA to identify differentially expressed genes (DEGs) within specific clusters. Additionally, we conducted GSVA to annotate enriched genes. Subsequently, we constructed and screened the best machine learning model. Finally, we validated the predictions' accuracy using a nomogram, calibration curves, decision curve analysis (DCA), and the GSE134431, GSE68183, and GSE80178 dataset.

RESULTS

In both the DFU and normal control groups, we confirmed the presence of deGlnMRGs and an activated immune response. From the GSE134431 dataset, we obtained 20 deGlnMRGs, including CTPS1, NAGS, SLC7A11, GGT1, GCLM, RIMKLA, ARG2, ASL, ASNS, ASNSD1, PPAT, GLS2, GLUD1, MECP2, ASS1, PRODH, CTPS2, ALDH5A1, DGLUCY, and SLC25A12. Furthermore, two clusters were identified in DFU. Immune infiltration analysis indicated the presence of immune heterogeneity in these two clusters. Additionally, we established a Support Vector Machine (SVM) model based on 5 genes (R3HCC1, ZNF562, MFN1, DRAM1, and PTGDS), which exhibited excellent performance on the external validation datasetGSE134431, GSE68183, and GSE80178 (AUC = 0.929).

CONCLUSION

This study has identified five Gln metabolism genes associated with DFU, revealing potential novel biomarkers and therapeutic targets for DFU. Additionally, the infiltration of immune-inflammatory cells plays a crucial role in the progression of DFU.

A robust framework for enhancing cardiovascular disease risk prediction using an optimized category boosting model

Cardiovascular disease (CVD) is a leading cause of mortality worldwide, and it is of utmost importance to accurately assess the risk of cardiovascular disease for prevention and intervention purposes. In recent years, machine learning has shown significant advancements in the field of cardiovascular disease risk prediction. In this context, we propose a novel framework known as CVD-OCSCatBoost, designed for the precise prediction of cardiovascular disease risk and the assessment of various risk factors. The framework utilizes Lasso regression for feature selection and incorporates an optimized category-boosting tree (CatBoost) model. Furthermore, we propose the opposition-based learning cuckoo search (OCS) algorithm. By integrating OCS with the CatBoost model, our objective is to develop OCSCatBoost, an enhanced classifier offering improved accuracy and efficiency in predicting CVD. Extensive comparisons with popular algorithms like the particle swarm optimization (PSO) algorithm, the seagull optimization algorithm (SOA), the cuckoo search algorithm (CS), K-nearest-neighbor classification, decision tree, logistic regression, grid-search support vector machine (SVM), grid-search XGBoost, default CatBoost, and grid-search CatBoost validate the efficacy of the OCSCatBoost algorithm. The experimental results demonstrate that the OCSCatBoost model achieves superior performance compared to other models, with overall accuracy, recall, and AUC values of 73.67%, 72.17%, and 0.8024, respectively. These outcomes highlight the potential of CVD-OCSCatBoost for improving cardiovascular disease risk prediction.

Predicting lncRNA-disease associations using multiple metapaths in hierarchical graph attention networks

BACKGROUND

Many biological studies have shown that lncRNAs regulate the expression of epigenetically related genes. The study of lncRNAs has helped to deepen our understanding of the pathogenesis of complex diseases at the molecular level. Due to the large number of lncRNAs and the complex and time-consuming nature of biological experiments, applying computer techniques to predict potential lncRNA-disease associations is very effective. To explore information between complex network structures, existing methods rely mainly on lncRNA and disease information. Metapaths have been applied to network models as an effective method for exploring information in heterogeneous graphs. However, existing methods are dominated by lncRNAs or disease nodes and tend to ignore the paths provided by intermediate nodes.

METHODS

We propose a deep learning model based on hierarchical graphical attention networks to predict unknown lncRNA-disease associations using multiple types of metapaths to extract features. We have named this model the MMHGAN. First, the model constructs a lncRNA-disease-miRNA heterogeneous graph based on known associations and two homogeneous graphs of lncRNAs and diseases. Second, for homogeneous graphs, the features of neighboring nodes are aggregated using a multihead attention mechanism. Third, for the heterogeneous graph, metapaths of different intermediate nodes are selected to construct subgraphs, and the importance of different types of metapaths is calculated and aggregated to obtain the final embedded features. Finally, the features are reconstructed using a fully connected layer to obtain the prediction results.

RESULTS

We used a fivefold cross-validation method and obtained an average AUC value of 96.07% and an average AUPR value of 93.23%. Additionally, ablation experiments demonstrated the role of homogeneous graphs and different intermediate node path weights. In addition, we studied lung cancer, esophageal carcinoma, and breast cancer. Among the 15 lncRNAs associated with these diseases, 15, 12, and 14 lncRNAs were validated by the lncRNA Disease Database and the Lnc2Cancer Database, respectively.

CONCLUSION

We compared the MMHGAN model with six existing models with better performance, and the case study demonstrated that the model was effective in predicting the correlation between potential lncRNAs and diseases.

Improving structure-based protein-ligand affinity prediction by graph representation learning and ensemble learning

Predicting protein-ligand binding affinity presents a viable solution for accelerating the discovery of new lead compounds. The recent widespread application of machine learning approaches, especially graph neural networks, has brought new advancements in this field. However, some existing structure-based methods treat protein macromolecules and ligand small molecules in the same way and ignore the data heterogeneity, potentially leading to incomplete exploration of the biochemical information of ligands. In this work, we propose LGN, a graph neural network-based fusion model with extra ligand feature extraction to effectively capture local features and global features within the protein-ligand complex, and make use of interaction fingerprints. By combining the ligand-based features and interaction fingerprints, LGN achieves Pearson correlation coefficients of up to 0.842 on the PDBbind 2016 core set, compared to 0.807 when using the features of complex graphs alone. Finally, we verify the rationalization and generalization of our model through comprehensive experiments. We also compare our model with state-of-the-art baseline methods, which validates the superiority of our model. To reduce the impact of data similarity, we increase the robustness of the model by incorporating ensemble learning.