Mendelian randomization highlights causal association between genetically increased C-reactive protein levels and reduced Alzheimer's disease risk

MGDDI: A multi-scale graph neural networks for drug-drug interaction prediction

Drug-drug interaction (DDI) prediction is crucial for identifying interactions within drug combinations, especially adverse effects due to physicochemical incompatibility. While current methods have made strides in predicting adverse drug interactions, limitations persist. Most methods rely on handcrafted features, restricting their applicability. They predominantly extract information from individual drugs, neglecting the importance of interaction details between drug pairs. To address these issues, we propose MGDDI, a graph neural network-based model for predicting potential adverse drug interactions. Notably, we use a multiscale graph neural network (MGNN) to learn drug molecule representations, addressing substructure size variations and preventing gradient issues. For capturing interaction details between drug pairs, we integrate a substructure interaction learning module based on attention mechanisms. Our experimental results demonstrate MGDDI's superiority in predicting adverse drug interactions, offering a solution to current methodological limitations.

Three-dimensional morphology scoring of hepatocellular carcinoma stratifies prognosis and immune infiltration

BACKGROUND

The morphological attributes could serve as pivotal indicators precipitating early recurrence and dismal overall survival in hepatocellular carcinoma (HCC), and quantifying morphological features may better stratify the prognosis of HCC.

OBJECTIVE

To develop a radiomics approach based on 3D tumor morphology features for predicting the prognosis of HCC and identifying differentially expressed genes related to morphology to guide HCC treatment.

MATERIALS AND METHODS

Retrospective study of 357 HCC patients. Radiomic features were extracted from MRI tumor regions; 14 morphology-related features predicted early HCC recurrence and patient stratification via LASSO-Cox modeling. Overall survival (OS) and recurrence-free survival (RFS) were analyzed. RNA sequencing from the Cancer Imaging Archive (TCIA) examined drug sensitivity and stratified HCC using morphological immunity genes, validating recurrence and prognosis.

RESULTS

Patients were split into training (n = 225), test (n = 132), and 50 TCIA dataset cohorts. Two features (Maximum2DdiameterColumn, Sphericity) in Cox regression stratified patients into high/low-risk Morphological Radiological Score (Morph-RS) groups. Significant OS and RFS were seen across all sets. Differentially expressed genes focused on T cell receptor signaling; low-risk group had higher T cells (P = 0.039), B cells (P = 0.041), NK cells (P = 0.018). SN-38, GSK2126458 might treat high-risk morphology. Morphology-immune genes stratified HCC, showing significant RFS/OS differences.

CONCLUSION

Tumor Morph-RS effectively stratifies HCC patients' recurrence and prognosis. Limited immune infiltration seen in Morph-RS high-risk groups signifies the potential of employing tumor morphology as a potent visual biomarker for diagnosing and managing HCC.

Fuzzy kernel evidence Random Forest for identifying pseudouridine sites

Pseudouridine is an RNA modification that is widely distributed in both prokaryotes and eukaryotes, and plays a critical role in numerous biological activities. Despite its importance, the precise identification of pseudouridine sites through experimental approaches poses significant challenges, requiring substantial time and resources.Therefore, there is a growing need for computational techniques that can reliably and quickly identify pseudouridine sites from vast amounts of RNA sequencing data. In this study, we propose fuzzy kernel evidence Random Forest (FKeERF) to identify pseudouridine sites. This method is called PseU-FKeERF, which demonstrates high accuracy in identifying pseudouridine sites from RNA sequencing data. The PseU-FKeERF model selected four RNA feature coding schemes with relatively good performance for feature combination, and then input them into the newly proposed FKeERF method for category prediction. FKeERF not only uses fuzzy logic to expand the original feature space, but also combines kernel methods that are easy to interpret in general for category prediction. Both cross-validation tests and independent tests on benchmark datasets have shown that PseU-FKeERF has better predictive performance than several state-of-the-art methods. This new method not only improves the accuracy of pseudouridine site identification, but also provides a certain reference for disease control and related drug development in the future.

Associations of environmental factors with neurodegeneration: An exposome-wide Mendelian randomization investigation

Neurodegenerative diseases (NDDs) remain a global health challenge. Previous studies have reported potential links between environmental factors and NDDs, however, findings remain controversial across studies and elusive to be interpreted as evidence of robust causal associations. In this study, we comprehensively explored the causal associations of the common environmental factors with major NDDs including Alzheimer's disease (AD), Parkinson's disease (PD), amyotrophic lateral sclerosis (ALS), and multiple sclerosis (MS), based on updated large-scale genome-wide association study data through two-sample Mendelian randomization (MR) approach. Our results indicated that, overall, 28 significant sets of exposure-outcome causal association evidence were detected, 12 of which were previously underestimated and newly identified, including average weekly beer plus cider intake, strenuous sports or other exercises, diastolic blood pressure, and body fat percentage with AD, alcohol intake frequency with PD, apolipoprotein B, systolic blood pressure, and forced expiratory volume in 1 s (FEV1) with ALS, and alcohol intake frequency, hip circumference, forced vital capacity, and FEV1 with MS. Moreover, the causal effects of several environmental factors on NDDs were found to overlap. From a triangulation perspective, our investigation provided insights into understanding the associations of environmental factors with NDDs, providing causality-oriented evidence to establish the risk profile of NDDs.

Genome-wide analysis reveals extensive genetic overlap between childhood phenotypes and later-life type 2 diabetes

Observational studies have indicated a potential influence of childhood phenotypes on the later development of type 2 diabetes (T2D). However, the underlying biological mechanisms remain unclear. In this study, we conducted a comprehensive genome-wide analysis to investigate the shared genetic architecture and genetic loci between nine childhood phenotypes (N = 4202-620,26) and later-life T2D (N = 80,154) using genetic correlation, mendelian randomization (MR), and conjunctional false discovery rate (conjFDR) statistical frameworks. Our findings demonstrated substantial genetic correlations and pleiotropic enrichment between childhood obesity, body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), and later-life T2D. Childhood obesity exhibited a significant association with increased later-life T2D risk through 10 mediators, 6 of which were adulthood obesity-related phenotypes. Additionally, we identified 69, 83, 3, 5, 10, 5, 3, and 7 loci shared between childhood obesity, BMI, SBP, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), apolipoprotein A-I (ApoA-I), apolipoprotein B (ApoB), and T2D at conjFDR <0.05, with the majority of these loci being novel discoveries. Overall, our study reveals extensive genetic overlap between childhood obesity-related phenotypes and T2D with concordant effect directions, shedding new light on variants and phenotypes with lifelong effects.

Mendelian randomization study supports the causal effects of air pollution on longevity via multiple age-related diseases

Growing evidence suggests that exposure to fine particulate matter (PM2.5) may reduce life expectancy; however, the causal pathways of PM2.5 exposure affecting life expectancy remain unknown. Here, we assess the causal effects of genetically predicted PM2.5 concentration on common chronic diseases and longevity using a Mendelian randomization (MR) statistical framework based on large-scale genome-wide association studies (GWAS) (>400,000 participants). After adjusting for other types of air pollution and smoking, we find significant causal relationships between PM2.5 concentration and angina pectoris, hypercholesterolaemia and hypothyroidism, but no causal relationship with longevity. Mediation analysis shows that although the association between PM2.5 concentration and longevity is not significant, PM2.5 exposure indirectly affects longevity via diastolic blood pressure (DBP), hypertension, angina pectoris, hypercholesterolaemia and Alzheimer's disease, with a mediated proportion of 31.5, 70.9, 2.5, 100, and 24.7%, respectively. Our findings indicate that public health policies to control air pollution may help improve life expectancy.

Prediction of thermophilic protein using 2-D general series correlation pseudo amino acid features

The demand for thermophilic protein has been increasing in protein engineering recently. Many machine-learning methods for identifying thermophilic proteins have emerged during this period. However, most machine learning-based thermophilic protein identification studies have only focused on accuracy. The relationship between the features' meaning and the proteins' physicochemical properties has yet to be studied in depth. In this article, we focused on the relationship between the features and the thermal stability of thermophilic proteins. This method used 2-D general series correlation pseudo amino acid (SC-PseAAC-General) features and realized accuracy of 82.76% using the J48 classifier. In addition, this research found the presence of higher frequencies of glutamic acid in thermophilic proteins, which help thermophilic proteins maintain their thermal stability by forming hydrogen bonds and salt bridges that prevent denaturation at high temperatures.

C-Reactive Protein: Pathophysiology, Diagnosis, False Test Results and a Novel Diagnostic Algorithm for Clinicians

The current literature provides a body of evidence on C-Reactive Protein (CRP) and its potential role in inflammation. However, most pieces of evidence are sparse and controversial. This critical state-of-the-art monography provides all the crucial data on the potential biochemical properties of the protein, along with further evidence on its potential pathobiology, both for its pentameric and monomeric forms, including information for its ligands as well as the possible function of autoantibodies against the protein. Furthermore, the current evidence on its potential utility as a biomarker of various diseases is presented, of all cardiovascular, respiratory, hepatobiliary, gastrointestinal, pancreatic, renal, gynecological, andrological, dental, oral, otorhinolaryngological, ophthalmological, dermatological, musculoskeletal, neurological, mental, splenic, thyroid conditions, as well as infections, autoimmune-supposed conditions and neoplasms, including other possible factors that have been linked with elevated concentrations of that protein. Moreover, data on molecular diagnostics on CRP are discussed, and possible etiologies of false test results are highlighted. Additionally, this review evaluates all current pieces of evidence on CRP and systemic inflammation, and highlights future goals. Finally, a novel diagnostic algorithm to carefully assess the CRP level for a precise diagnosis of a medical condition is illustrated.

Comparative evaluation and analysis of DNA N4-methylcytosine methylation sites using deep learning

DNA N4-methylcytosine (4mC) is significantly involved in biological processes, such as DNA expression, repair, and replication. Therefore, accurate prediction methods are urgently needed. Deep learning methods have transformed applications that previously require sequencing expertise into engineering challenges that do not require expertise to solve. Here, we compare a variety of state-of-the-art deep learning models on six benchmark datasets to evaluate their performance in 4mC methylation site detection. We visualize the statistical analysis of the datasets and the performance of different deep-learning models. We conclude that deep learning can greatly expand the potential of methylation site prediction.

A First Computational Frame for Recognizing Heparin-Binding Protein

Heparin-binding protein (HBP) is a cationic antibacterial protein derived from multinuclear neutrophils and an important biomarker of infectious diseases. The correct identification of HBP is of great significance to the study of infectious diseases. This work provides the first HBP recognition framework based on machine learning to accurately identify HBP. By using four sequence descriptors, HBP and non-HBP samples were represented by discrete numbers. By inputting these features into a support vector machine (SVM) and random forest (RF) algorithm and comparing the prediction performances of these methods on training data and independent test data, it is found that the SVM-based classifier has the greatest potential to identify HBP. The model could produce an auROC of 0.981 ± 0.028 on training data using 10-fold cross-validation and an overall accuracy of 95.0% on independent test data. As the first model for HBP recognition, it will provide some help for infectious diseases and stimulate further research in related fields.

Computational identification of promoters in Klebsiella aerogenes by using support vector machine

Promoters are the basic functional cis-elements to which RNA polymerase binds to initiate the process of gene transcription. Comprehensive understanding gene expression and regulation depends on the precise identification of promoters, as they are the most important component of gene expression. This study aimed to develop a machine learning-based model to predict promoters in Klebsiella aerogenes (K. aerogenes). In the prediction model, the promoter sequences in K. aerogenes genome were encoded by pseudo k-tuple nucleotide composition (PseKNC) and position-correlation scoring function (PCSF). Numerical features were obtained and then optimized using mRMR by combining with support vector machine (SVM) and 5-fold cross-validation (CV). Subsequently, these optimized features were inputted into SVM-based classifier to discriminate promoter sequences from non-promoter sequences in K. aerogenes. Results of 10-fold CV showed that the model could yield the overall accuracy of 96.0% and the area under the ROC curve (AUC) of 0.990. We hope that this model will provide help for the study of promoter and gene regulation in K. aerogenes.

Empirical Comparison and Recent Advances of Computational Prediction of Hormone Binding Proteins Using Machine Learning Methods

Hormone binding proteins (HBPs) belong to the group of soluble carrier proteins. These proteins selectively and non-covalently interact with hormones and promote growth hormone signaling in human and other animals. The HBPs are useful in many medical and commercial fields. Thus, the identification of HBPs is very important because it can help to discover more details about hormone binding proteins. Meanwhile, the experimental methods are time-consuming and expensive for hormone binding proteins recognition. Computational prediction methods have played significant roles in the correct recognition of hormone binding proteins with the use of sequence information and ML algorithms. In this review, we compared and assessed the implementation of ML-based tools in recognition of HBPs in a unique way. We hope that this study will give enough awareness and knowledge for research on HBPs.

KRAS is a prognostic biomarker associated with diagnosis and treatment in multiple cancers

KRAS encodes K-Ras proteins, which take part in the MAPK pathway. The expression level of KRAS is high in tumor patients. Our study compared KRAS expression levels between 33 kinds of tumor tissues. Additionally, we studied the association of KRAS expression levels with diagnostic and prognostic values, clinicopathological features, and tumor immunity. We established 22 immune-infiltrating cell expression datasets to calculate immune and stromal scores to evaluate the tumor microenvironment. KRAS genes, immune check-point genes and interacting genes were selected to construct the PPI network. We selected 79 immune checkpoint genes and interacting related genes to calculate the correlation. Based on the 33 tumor expression datasets, we conducted GSEA (genome set enrichment analysis) to show the KRAS and other co-expressed genes associated with cancers. KRAS may be a reliable prognostic biomarker in the diagnosis of cancer patients and has the potential to be included in cancer-targeted drugs.

MultiscaleDTA: a multiscale-based method with a self-attention mechanism for drug-target binding affinity prediction

The task of predicting drug-target affinity (DTA) plays an increasingly important role at the early stage of in silico drug discovery and development. Currently, a variety of machine learning-based methods have been presented for DTA prediction and achieved outstanding performance, which is beneficial for speeding up the development of new drugs. However, most convolutional neural networks (CNNs) based methods ignore the significance of information from CNN layers with different scales to DTA prediction. In addition, each feature provides different contributions to the final task. Therefore, in this study, we propose a novel end-to-end deep learning-based framework, called MultiscaleDTA, to predict drug-target binding affinity. MultiscaleDTA incorporates multi-scale CNNs and a self-attention mechanism to capture multi-scale and comprehensive features for characterizing the intrinsic properties of drugs and targets. Extensive experimental results on both regression and binary classification tasks demonstrate that MultiscaleDTA has achieved competitive performance compared to state-of-the-art methods.

Mendelian randomization analysis of the causal association of bone mineral density and fracture with multiple sclerosis

Multiple sclerosis (MS) is a neurodegenerative disorder and an autoimmune disease. Until now, observational studies have indicated the association of bone mineral density (BMD) and fracture with the risk of MS. However, these studies indicated inconsistent findings. Until now, genome-wide association studies (GWAS) have been conducted in BMD, fracture, and MS, which provide large-scale datasets to investigate the causal association of BMD and fracture with the risk of MS using the Mendelian randomization (MR) study. Here, we performed an MR study to clarify the causal association between BMD/fracture and the risk of MS using large-scale publicly available GWAS datasets from BMD, fracture, and MS. We first evaluated the bidirectional causal effects of BMD and MS. The main analysis method inverse-variance weighted (IVW) showed no significant causal effect of BMD on the risk of MS (β = 0.058, and p = 1.98E-01), and MS on the risk of BMD (β = −0.001, and p = 7.83E-01). We then evaluated the bidirectional causal effects of fracture and MS. However, we only identified a significant causal effect of fracture on the risk of MS using IVW (β = −0.375, p = 0.002), but no significant causal effect of MS on the risk of the fracture using IVW (β = 0.011, p = 2.39E-01). Therefore, our main analysis method IVW only found a significant causal effect of fracture on MS using the threshold for the statistically significant association p < 0.05/4 = 0.0125. Meanwhile, multivariable MR analyses showed that the causal effect of fracture on MS was independent of smoking, drinking, and obesity, but dependent on BMD. In summary, our MR analysis demonstrates that genetically increased fracture may reduce the risk of MS. Our findings should be further verified and the underlying mechanisms should be further evaluated by future studies.

Evaluating the Bidirectional Causal Association Between Daytime Napping and Alzheimer’s Disease Using Mendelian Randomization

Background: Until now, both cross-sectional and longitudinal studies have identified controversial findings about the association between daytime napping and Alzheimer’s disease (AD) or cognitive decline. Therefore, it remains unclear about the causal association between daytime napping and AD or cognitive decline. Objective: We aim to investigate the causal association between daytime napping and AD. Methods: Here, we conduct a bidirectional Mendelian randomization (MR) analysis to investigate the causal association between daytime napping and AD using large-scale GWAS datasets from daytime napping including 452,633 individuals of European ancestry and AD including 35,274 AD and 59,163 controls of European ancestry. A total of five MR methods are selected including inverse-variance weighted (IVW), weighted median, MR-Egger, MR-PRESSO, and contamination mixture method. Results: MR analysis highlights significant causal association of AD with daytime napping using IVW (beta = -0.006, 95% CI [–0.009, –0.002], p = 2.00E-03), but no significant causal association of daytime napping with AD using IVW (OR = 0.76, 95% CI 0.53-1.10, p = 1.40E-01). Conclusion: Our bidirectional MR analysis demonstrates the causal effect of AD on daytime napping. However, there is no causal effect of daytime napping on AD. Our current findings are consistent with recent evidence from other MR studies that highlight little evidence supporting a causal effect of sleep traits on AD and support the causal effect of AD on sleep traits.