MDSVDNV: predicting microbe-drug associations by singular value decomposition and Node2vec

Introduction

Recent researches have demonstrated that microbes are crucial for the growth and development of the human body, the movement of nutrients, and human health. Diseases may arise as a result of disruptions and imbalances in the microbiome. The pathological investigation of associated diseases and the advancement of clinical medicine can both benefit from the identification of drug-associated microbes.

Methods

In this article, we proposed a new prediction model called MDSVDNV to infer potential microbe-drug associations, in which the Node2vec network embedding approach and the singular value decomposition (SVD) matrix decomposition method were first adopted to produce linear and non-linear representations of microbe interactions.

Results and discussion

Compared with state-of-the-art competitive methods, intensive experimental results demonstrated that MDSVDNV could achieve the best AUC value of 98.51% under a 5-fold CV, which indicated that MDSVDNV outperformed existing competing models and may be an effective method for discovering latent microbe-drug associations in the future.

Multi-Class Seizure Type Classification Using Features Extracted from the EEG

In this study, we classify the seizure types using feature extraction and machine learning algorithms. Initially, we pre-processed the electroencephalogram (EEG) of focal non-specific seizure (FNSZ), generalized seizure (GNSZ), tonic-clonic seizure (TCSZ), complex partial seizure (CPSZ) and absence seizure (ABSZ). Further, 21 features from time (9) and frequency (12) domain were computed from the EEG signals of different seizure types. XGBoost classifier model was built for individual domain features and combination of time and frequency features and validated the results using 10-fold cross-validation. Our results revealed that the classifier model with combination of time and frequency features performed well followed by the time and frequency domain features. We obtained a highest multi-class accuracy of 79.72% for the classification of five types of seizure while using all the 21 features. The band power between 11-13 Hz was found to be the top feature in our study. The proposed study can be used for the seizure type classification in clinical applications.

Using Deep Learning Architectures for Detection and Classification of Diabetic Retinopathy

Diabetic retinopathy (DR) is a common complication of long-term diabetes, affecting the human eye and potentially leading to permanent blindness. The early detection of DR is crucial for effective treatment, as symptoms often manifest in later stages. The manual grading of retinal images is time-consuming, prone to errors, and lacks patient-friendliness. In this study, we propose two deep learning (DL) architectures, a hybrid network combining VGG16 and XGBoost Classifier, and the DenseNet 121 network, for DR detection and classification. To evaluate the two DL models, we preprocessed a collection of retinal images obtained from the APTOS 2019 Blindness Detection Kaggle Dataset. This dataset exhibits an imbalanced image class distribution, which we addressed through appropriate balancing techniques. The performance of the considered models was assessed in terms of accuracy. The results showed that the hybrid network achieved an accuracy of 79.50%, while the DenseNet 121 model achieved an accuracy of 97.30%. Furthermore, a comparative analysis with existing methods utilizing the same dataset revealed the superior performance of the DenseNet 121 network. The findings of this study demonstrate the potential of DL architectures for the early detection and classification of DR. The superior performance of the DenseNet 121 model highlights its effectiveness in this domain. The implementation of such automated methods can significantly improve the efficiency and accuracy of DR diagnosis, benefiting both healthcare providers and patients.

LDAEXC: LncRNA-Disease Associations Prediction with Deep Autoencoder and XGBoost Classifier

Numerous scientific evidences have revealed that long non-coding RNAs (lncRNAs) are involved in the progression of human complex diseases and biological life activities. Therefore, identifying novel and potential disease-related lncRNAs is helpful to diagnosis, prognosis and therapy of many human complex diseases. Since traditional laboratory experiments are cost and time-consuming, a great quantity of computer algorithms have been proposed for predicting the relationships between lncRNAs and diseases. However, there are still much room for the improvement. In this paper, we introduce an accurate framework named LDAEXC to infer LncRNA-Disease Associations with deep autoencoder and XGBoost Classifier. LDAEXC utilizes different similarity views of lncRNAs and human diseases to construct features for each data sources. Then, the reduced features are obtained by feeding the constructed feature vectors into a deep autoencoder, and at last an XGBoost classifier is leveraged to calculate the latent lncRNA-disease-associated scores using reduced features. The fivefold cross-validation experiments on four datasets showed that LDAEXC reached AUC scores of 0.9676 ± 0.0043, 0.9449 ± 0.022, 0.9375 ± 0.0331 and 0.9556 ± 0.0134, respectively, significantly higher than other advanced similar computer methods. Extensive experiment results and case studies of two complex diseases (colon and breast cancers) further indicated the practicability and excellent prediction performance of LDAEXC in inferring unknown lncRNA-disease associations. TLDAEXC utilizes disease semantic similarity, lncRNA expression similarity, and Gaussian interaction profile kernel similarity of lncRNAs and diseases for feature construction. The constructed features are fed to a deep autoencoder to extract reduced features, and an XGBoost classifier is used to predict the lncRNA-disease associations based on the reduced features. The fivefold and tenfold cross-validation experiments on a benchmark dataset showed that LDAEXC could achieve AUC scores of 0.9676 and 0.9682, respectively, significantly higher than other state-of-the-art similar methods.

A random forest-based metabolic risk model to assess the prognosis and metabolism-related drug targets in ovarian cancer

As one of the most common gynecologic malignant tumors, ovarian cancer is usually diagnosed at an advanced and incurable stage because of its early asymptomatic onset. Increasing research into tumor biology has demonstrated that abnormal cellular metabolism precedes tumorigenesis, therefore it has become an area of active research in academia. Cellular metabolism is of great significance in cancer diagnostic and prognostic studies. In this study, we integrated The Cancer Genome Atlas dataset with multiple Gene Expression Omnibus ovarian cancer datasets, identified 17 metabolic pathways with prognostic values using the random forest algorithm, constructed a metabolic risk scoring model based on metabolic pathway enrichment scores, and classified patients with ovarian cancer into two subtypes. Then, we systematically investigated the differences between different subtypes in terms of prognosis, differential gene expression, immune signature enrichment, Hallmark signature enrichment, and somatic mutations. As well, we successfully predicted differences in sensitivity to immunotherapy and chemotherapy drugs in patients with different metabolic risk subtypes. Moreover, we identified 5 drug targets associated with high metabolic risk and low metabolic risk ovarian cancer phenotypes through the weighted correlation network analysis and investigated their roles in the genesis of ovarian cancer. Finally, we developed an XGBoost classifier for predicting metabolic risk types in patients with ovarian cancer, producing a good predictive effect. In light of the above study, the research findings will provide valuable information for prognostic prediction and personalized medical treatment of patients with ovarian cancer.

The Effect of Primary Aldosteronism on Carotid Artery Texture in Ultrasound Images

Primary aldosteronism (PA) is the most frequent cause of secondary hypertension. Early diagnoses of PA are essential to avoid the long-term negative effects of elevated aldosterone concentration on the cardiovascular and renal system. In this work, we study the texture of the carotid artery vessel wall from longitudinal ultrasound images in order to automatically distinguish between PA and essential hypertension (EH). The texture is characterized using 140 Haralick and 10 wavelet features evaluated in a region of interest in the vessel wall, followed by the XGBoost classifier. Carotid ultrasound studies were carried out on 33 patients aged 42-72 years with PA, 52 patients with EH, and 33 normotensive controls. For the most clinically relevant task of distinguishing PA and EH classes, we achieved a classification accuracy of 73% as assessed by a leave-one-out procedure. This result is promising even compared to the 57% prediction accuracy using clinical characteristics alone or 63% accuracy using a combination of clinical characteristics and intima-media thickness (IMT) parameters. If the accuracy is improved and the method incorporated into standard clinical procedures, this could eventually lead to an improvement in the early diagnosis of PA and consequently improve the clinical outcome for these patients in future.

Using ECG signals for hypotensive episode prediction in trauma patients

BACKGROUND AND OBJECTIVES

Bleeding is the leading cause of death among trauma patients both in military and civilian scenarios, and it is also the most common cause of preventable death. Identifying a casualty who suffers from an internal bleeding and may deteriorate rapidly and develop hemorrhagic shock and multiorgan failure is a profound challenge. Blood pressure and heart rate are the main vital signs used nowadays for the casualty clinical evaluation in the battlefield and in intensive care unit. However, these vitals tend to deteriorate at a relatively late stage, when the ability to prevent hazardous complications is limited. Identifying, treating, and rapidly evacuating such casualties might mitigate these complications. In this work, we try to improve a state-of-the-art method for early identification of Hypotensive Episode (HE), by adding electrocardiogram signals to several vital signs.

METHODS

In this research, we propose to extend the state-of-the-art HE early detection method, In-Window Segmentation (InWise), by adding new types of features extracted from ECG signals. The new predictive features can be extracted from ECG signals both manually and automatically by a convolutional auto-encoder. In addition to InWise, we are trying to predict HE using a Transformer model. The Transformer is using the encoder output as an embedding of the ECG signal. The proposed approach is evaluated on trauma patients data from the MIMIC III database.

RESULTS

We evaluated the InWise prediction algorithm using four different groups of features. The first feature group contains the 93 original features extracted from vital signs. The second group contains, in addition to the original features, 24 features extracted manually from ECG signal (117 features in total). The third group contains the original features and 20 ECG features extracted by the AE (113 features in total), and the last group is the union of all three previous groups containing 137 features. The results show that each model, which has used ECG data, is outperforming the original InWise model, in terms of AUC and sensitivity with p-value <0.001 (by 0.7% in AUC and up to 3.8% in sensitivity). The model which has used all three feature types (vital signs, manual ECG and AE ECG), outperforms the original model both in terms of accuracy and specificity with p-value <0.001 (by 0.3% and 0.4% respectively).

CONCLUSION

The results show an improvement in the prediction success rates as a result of using ECG-based features. The importance of ECG features was confirmed by the feature importance analysis.

SVDNVLDA: predicting lncRNA-disease associations by Singular Value Decomposition and node2vec

Background

Numerous studies on discovering the roles of long non-coding RNAs (lncRNAs) in the occurrence, development and prognosis progresses of various human diseases have drawn substantial attentions. Since only a tiny portion of lncRNA-disease associations have been properly annotated, an increasing number of computational methods have been proposed for predicting potential lncRNA-disease associations. However, traditional predicting models lack the ability to precisely extract features of biomolecules, it is urgent to find a model which can identify potential lncRNA-disease associations with both efficiency and accuracy.

Results

In this study, we proposed a novel model, SVDNVLDA, which gained the linear and non-linear features of lncRNAs and diseases with Singular Value Decomposition (SVD) and node2vec methods respectively. The integrated features were constructed from connecting the linear and non-linear features of each entity, which could effectively enhance the semantics contained in ultimate representations. And an XGBoost classifier was employed for identifying potential lncRNA-disease associations eventually.

Conclusions

We propose a novel model to predict lncRNA-disease associations. This model is expected to identify potential relationships between lncRNAs and diseases and further explore the disease mechanisms at the lncRNA molecular level.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04457-1.

Using Vehicle Interior Noise Classification for Monitoring Urban Rail Transit Infrastructure

This study developed a multi-classification model for vehicle interior noise from the subway system, collected on smartphones. The proposed model has the potential to be used to analyze the causes of abnormal noise using statistical methods and evaluate the effect of rail maintenance work. To this end, first, we developed a multi-source data (audio, acceleration, and angle rate) collection framework via smartphone built-in sensors. Then, considering the Shannon entropy, a 1-second window was selected to segment the time-series signals. This study extracted 45 features from the time- and frequency-domains to establish the classifier. Next, we investigated the effects of balancing the training dataset with the Synthetic Minority Oversampling Technique (SMOTE). By comparing and analyzing the classification results of importance-based and mutual information-based feature selection methods, the study employed a feature set consisting of the top 10 features by importance score. Comparisons with other classifiers indicated that the proposed XGBoost-based classifier runs fast while maintaining good accuracy. Finally, case studies were provided to extend the applications of this classifier to the analysis of abnormal vehicle interior noise events and evaluate the effects of rail grinding.

Prediction of drug-target interaction based on protein features using undersampling and feature selection techniques with boosting

Accurate identification of drug-target interaction (DTI) is a crucial and challenging task in the drug discovery process, having enormous benefit to the patients and pharmaceutical company. The traditional wet-lab experiments of DTI is expensive, time-consuming, and labor-intensive. Therefore, many computational techniques have been established for this purpose; although a huge number of interactions are still undiscovered. Here, we present pdti-EssB, a new computational model for identification of DTI using protein sequence and drug molecular structure. More specifically, each drug molecule is transformed as the molecular substructure fingerprint. For a protein sequence, different descriptors are utilized to represent its evolutionary, sequence, and structural information. Besides, our proposed method uses data balancing techniques to handle the imbalance problem and applies a novel feature eliminator to extract the best optimal features for accurate prediction. In this paper, four classes of DTI benchmark datasets are used to construct a predictive model with XGBoost. Here, the auROC is utilized as an evaluation metric to compare the performance of pdti-EssB method with recent methods, applying five-fold cross-validation. Finally, the experimental results indicate that our proposed method is able to outperform other approaches in predicting DTI, and introduces new drug-target interaction samples based on prediction probability scores. pdti-EssB webserver is available online at http://pdtiessb-uestc.com/.

Five-Feature Model for Developing the Classifier for Synergistic vs. Antagonistic Drug Combinations Built by XGBoost

Combinatorial drug therapy can improve the therapeutic effect and reduce the corresponding adverse events. In silico strategies to classify synergistic vs. antagonistic drug pairs is more efficient than experimental strategies. However, most of the developed methods have been applied only to cancer therapies. In this study, we introduce a novel method, XGBoost, based on five features of drugs and biomolecular networks of their targets, to classify synergistic vs. antagonistic drug combinations from different drug categories. We found that XGBoost outperformed other classifiers in both stratified fivefold cross-validation (CV) and independent validation. For example, XGBoost achieved higher predictive accuracy than other models (0.86, 0.78, 0.78, and 0.83 for XGBoost, logistic regression, naïve Bayesian, and random forest, respectively) for an independent validation set. We also found that the five-feature XGBoost model is much more effective at predicting combinatorial therapies that have synergistic effects than those with antagonistic effects. The five-feature XGBoost model was also validated on TCGA data with accuracy of 0.79 among the 61 tested drug pairs, which is comparable to that of DeepSynergy. Among the 14 main anatomical/pharmacological groups classified according to WHO Anatomic Therapeutic Class, for drugs belonging to five groups, their prediction accuracy was significantly increased (odds ratio < 1) or reduced (odds ratio > 1) (Fisher's exact test, p < 0.05). This study concludes that our five-feature XGBoost model has significant benefits for classifying synergistic vs. antagonistic drug combinations.