Search

Motivation 5-Methylcytosine (m5C) is a crucial post-transcriptional modification. With the development of technology, it is widely found in various RNAs. Numerous studies have indicated that m5C plays an essential role in various activities of organisms, such as tRNA recognition, stabilization of RNA structure, RNA metabolism and so on. Traditional identification is costly and time-consuming by wet biological experiments. Therefore, computational models are commonly used to identify the m5C sites. Due to the vast computing advantages of deep learning, it is feasible to construct the predictive model through deep learning algorithms. Results In this study, we construct a model to identify m5C based on a deep fusion approach with an improved residual network. First, sequence features are extracted from the RNA sequences using Kmer, K-tuple nucleotide frequency component (KNFC), Pseudo dinucleotide composition (PseDNC) and Physical and chemical property (PCP). Kmer and KNFC extract information from a statistical point of view. PseDNC and PCP extract information from the physicochemical properties of RNA sequences. Then, two parts of information are fused with new features using bidirectional long- and short-term memory and attention mechanisms, respectively. Immediately after, the fused features are fed into the improved residual network for classification. Finally, 10-fold cross-validation and independent set testing are used to verify the credibility of the model. The results show that the accuracy reaches 91.87%, 95.55%, 92.27% and 95.60% on the training sets and independent test sets of Arabidopsis thaliana and M.musculus , respectively. This is a considerable improvement compared to previous studies and demonstrates the robust performance of our model. Availability and implementation The data and code related to the study are available at https://github.com/alivelxj/m5c-DFRESG. [ABSTRACT FROM AUTHOR]

RNA 5-hydroxymethylcytosine (5hmC) is a kind of RNA modification, which is related to the life activities of many organisms. Studying its distribution is very important to reveal its biological function. Previously, high-throughput sequencing was used to identify 5hmC, but it is expensive and inefficient. Therefore, machine learning is used to identify 5hmC sites. Here, we design a model called R5hmCFDV, which is mainly divided into feature representation, feature fusion and classification. (i) Pseudo dinucleotide composition, dinucleotide binary profile and frequency, natural vector and physicochemical property are used to extract features from four aspects: nucleotide composition, coding, natural language and physical and chemical properties. (ii) To strengthen the relevance of features, we construct a novel feature fusion method. Firstly, the attention mechanism is employed to process four single features, stitch them together and feed them to the convolution layer. After that, the output data are processed by BiGRU and BiLSTM, respectively. Finally, the features of these two parts are fused by the multiply function. (iii) We design the deep voting algorithm for classification by imitating the soft voting mechanism in the Python package. The base classifiers contain deep neural network (DNN), convolutional neural network (CNN) and improved gated recurrent unit (GRU). And then using the principle of soft voting, the corresponding weights are assigned to the predicted probabilities of the three classifiers. The predicted probability values are multiplied by the corresponding weights and then summed to obtain the final prediction results. We use 10-fold cross-validation to evaluate the model, and the evaluation indicators are significantly improved. The prediction accuracy of the two datasets is as high as 95.41% and 93.50%, respectively. It demonstrates the stronger competitiveness and generalization performance of our model. In addition, all datasets and source codes can be found at https://github.com/HongyanShi026/R5hmCFDV. [ABSTRACT FROM AUTHOR]

Although deep convolutional neural networks have achieved remarkable success in removing synthetic fog, it is essential to be able to process images taken in complex foggy conditions, such as dense or non-homogeneous fog, in the real world. However, the haze distribution in the real world is complex, and downsampling can lead to color distortion or loss of detail in the output results as the resolution of a feature map or image resolution decreases. Moreover, the over-stacking of convolutional blocks might increase the model complexity. In addition to the challenges of obtaining sufficient training data, overfitting can also arise in deep learning techniques for foggy image processing, which can limit the generalization abilities of the model, posing challenges for its practical applications in real-world scenarios. Considering these issues, this paper proposes a Transformer-based wavelet network (WaveletFormerNet) for real-world foggy image recovery. We embed the discrete wavelet transform into the Vision Transformer by proposing the WaveletFormer and IWaveletFormer blocks, aiming to alleviate texture detail loss and color distortion in the image due to downsampling. We introduce parallel convolution in the Transformer block, which allows for the capture of multi-frequency information in a lightweight mechanism. Such a structure reduces computational expenses and improves the effectiveness of the network. Additionally, we have implemented a feature aggregation module (FAM) to maintain image resolution and enhance the feature extraction capacity of our model, further contributing to its impressive performance in real-world foggy image recovery tasks. Through extensive experiments on real-world fog datasets, we have demonstrated that our WaveletFormerNet achieves superior performance compared to state-of-the-art methods, as shown through quantitative and qualitative evaluations of minor model complexity. Additionally, our satisfactory results on real-world dust removal and application tests showcase the superior generalization ability and improved performance of WaveletFormerNet in computer vision-related applications compared to existing state-of-the-art methods, further confirming our proposed approach's effectiveness and robustness. Our code is available at https://github.com/shengli666666/WaveletFormerNet. [Display omitted] • Combining Vision Transformer and wavelet transform alleviates image detail loss. • Parallel convolution captures frequency information in a lightweight mechanism. • The feature aggregation module enhances the model feature extraction capability. • Experimental results demonstrate WaveletFormerNet's superior dehazing performance. [ABSTRACT FROM AUTHOR]

DNA-binding proteins play a vital role in cellular processes. It is an extremely urgent to develop a high-throughput method for efficiently identifying DNA-binding proteins. According to the current research situation, some methods in machine learning and deep learning show excellent computational speed and accuracy, which are worthy of application. In this work, a novel predictor was proposed to predict DNA binding proteins called UMAP-DBP. Firstly, the feature extraction of primary protein sequence was realized based on physicochemical distance transformation, Profile-based auto-cross covariance and General series correlation pseudo amino acid composition. Secondly, uniform manifold approximation and projection (UMAP) and feature importance score methods were used for feature selection; there is a progressive relationship between them. Finally, the Adaboost operation engine with jackknife test were adopted for predicting DNA-binding proteins. For the jackknife test on the BP1075 and BP594, we obtained an overall accuracy of 82.97% and 82.14%, Cohen's kappa (CK) of 0.66 and 0.64, respectively. The results illustrate that a feasible method has been developed for predicting DNA-binding proteins by UMAP and Adaboost. This is the first study in which UMAP has been successfully applied to identify DNA-binding proteins. All the datasets and codes are accessible at https://github.com/Wang-Jinyue/UMAP-DBP. [ABSTRACT FROM AUTHOR]

Meiotic recombination plays an important role in the process of genetic evolution. Previous researches have shown that the recombination rates provide important information about the mechanism of recombination study. However, at present, most methods ignore the hidden correlation and spatial autocorrelation of the DNA sequence. In this study, we proposed a predictor called iRSpot-DTS to identify hot/cold spots based on the benchmark datasets. We proposed a feature extraction method called dinucleotide-based spatial autocorrelation(DSA) which can incorporate the original DNA properties and spatial information of DNA sequence. Then it used t-SNE method to remove the noise which outperformed PCA. Finally, we used SAE softmax classifier to do classification which is based on networks and can get more hidden information of DNA sequence, our iRSpot-DTS achieved remarkable performance. Jackknife cross validation tests were done on two benchmark datasets. We achieved state-of-the-art results with 96.61% overall accuracy(OA), 93.16% Matthews correlation coefficient (MCC) and over 95% in Sn and Sp which are the best in this state. A predictor called iRSpot-DTS to identify hot/cold spots based on the benchmark datasets. The DSA method can incorporate the original DNA properties and spatial information of DNA sequence. It used t-SNE method to remove the noise and used SAE softmax classifier to do classification which is based on networks. The iRSpot-DTS model achieves the remarkable performance. [ABSTRACT FROM AUTHOR]

Gear fault diagnosis relies heavily on the scrutiny of vibration responses measured. In reality, gear vibration signals are noisy and dominated by meshing frequencies as well as their harmonics, which oftentimes overlay the fault related components. Moreover, many gear transmission systems, e.g., those in wind turbines, constantly operate under non-stationary conditions. To reduce the influences of non-synchronous components and noise, a fault signature enhancement method that is built upon angle-frequency domain synchronous averaging is developed in this paper. Instead of being averaged in the time domain, the signals are processed in the angle-frequency domain to solve the issue of phase shifts between signal segments due to uncertainties caused by clearances, input disturbances, and sampling errors, etc. The enhanced results are then analyzed through feature extraction algorithms to identify the most distinct features for fault classification and identification. Specifically, Kernel Principal Component Analysis (KPCA) targeting at nonlinearity, Multilinear Principal Component Analysis (MPCA) targeting at high dimensionality, and Locally Linear Embedding (LLE) targeting at local similarity among the enhanced data are employed and compared to yield insights. Numerical and experimental investigations are performed, and the results reveal the effectiveness of angle-frequency domain synchronous averaging in enabling feature extraction and classification. [ABSTRACT FROM AUTHOR]

Structural class is an important attribute of protein, which plays an important role in both the experiment and theoretical research of protein science. Prediction of protein structural classes has been a challenging task, specifically for low-similarity datasets. In this paper, we develop a feature extraction method PSSS-SOMA-PSSM by incorporating two different modes of evolutionary information into Chou's general pseudo amino acid composition (PseAAC) based on predicted secondary structure sequence (PSSS) and position-specific scoring matrix (PSSM). We construct a 170-dimensional (170D) feature vector for each protein sequence sample, which contains 10D PSSS features that reflect content, alternating word frequency and novel position information, and contains 160D PSSM features that are calculated by second-order moving average (SOMA) algorithm. The SVM classifier with RBF kernel function and the jackknife test are used to predict and evaluate on 1189, 25PDB and 640 datasets with sequence similarity lower than 40%, 25%, and 25%, respectively. Comparison of our results with other methods shows that the proposed method provides the state-of-the-art performance and a cost-effective alternative to structural classes prediction for low-similarity datasets. [ABSTRACT FROM AUTHOR]

Human leukocyte antigen (HLA) plays a vital role in immunomodulatory function. Studies have shown that immunotherapy based on non-classical HLA has essential applications in cancer, COVID-19, and allergic diseases. However, there are few deep learning methods to predict non-classical HLA alleles. In this work, an adaptive dual-attention network named DapNet-HLA is established based on existing datasets. Firstly, amino acid sequences are transformed into digital vectors by looking up the table. To overcome the feature sparsity problem caused by unique one-hot encoding, the fused word embedding method is used to map each amino acid to a low-dimensional word vector optimized with the training of the classifier. Then, we use the GCB (group convolution block), SENet attention (squeeze-and-excitation networks), BiLSTM (bidirectional long short-term memory network), and Bahdanau attention mechanism to construct the classifier. The use of SENet can make the weight of the effective feature map high, so that the model can be trained to achieve better results. Attention mechanism is an Encoder-Decoder model used to improve the effectiveness of RNN, LSTM or GRU (gated recurrent neural network). The ablation experiment shows that DapNet-HLA has the best adaptability for five datasets. On the five test datasets, the ACC index and MCC index of DapNet-HLA are 4.89% and 0.0933 higher than the comparison method, respectively. According to the ROC curve and PR curve verified by the 5-fold cross-validation, the AUC value of each fold has a slight fluctuation, which proves the robustness of the DapNet-HLA. The codes and datasets are accessible at https://github.com/JYY625/DapNet-HLA. [Display omitted] • A novel model named DapNet-HLA is proposed to predict non-classical HLA binding sites. • Fused word embedding methods are applied to extract features from the datasets. • SENet and Bahdanau attention mechanisms are integrated to optimize and improve the effectiveness of BiLSTM. • Adaptive dual-attention mechanism network based on deep learning is designed to predict non-classical HLA binding sites. [ABSTRACT FROM AUTHOR]

• A novel predictor called iPro-GAN to identify promoters and their strength. • Moran-based spatial auto-cross correlation considering physicochemical properties of dinucleotide is used to extract sequence features. • A generative adversarial network based on deep convolution is established to solve the binary classification problem in bioinformatics for the first time. Promoter is a component of the gene, which can specifically bind with RNA polymerase and determine where transcription starts, and also determine the transcription efficiency of the gene. Promoters can be divided into strong promoters and weak promoters because their structures and the interaction time interval are quite different. The functional variation of the promoter can lead to a variety of diseases. Therefore, identifying promoters and their strength is necessary and has important biological significance. A novel and promising model based on deep learning is proposed to achieve it. In this work, we build a power model named iPro-GAN for identification of promoters and their strength. First, we collect benchmark datasets and independent datasets for training and testing. Then, Moran-based spatial auto-cross correlation method is used as feature extraction method. Finally, deep convolution generative adversarial network with 10-fold cross validation is applied for classifying. The first layer of the model is used to identify the promoter and the second layer is used to determine its type. On the benchmark data set, the accuracy of the first layer predictor is 93.15%, and the accuracy of the second layer predictor is 92.30%. On the independent data set, the accuracy of the first layer predictor is 86.77%, and the accuracy of the second layer predictor is 91.66%. In particular, breakthrough progress has been made in the identification of promoters' strength. These results are far higher than the existing best predictor, which indicate that our model is serviceable and practicable to identify promoters and their strength. Furthermore, the datasets and source codes are available from this link: https://github.com/Bovbene/iPro-GAN. [ABSTRACT FROM AUTHOR]

Therapeutic peptides, as active substances involved in a variety of cell functions in the organism, are essential participants to complete complex physiological activities of the body. Therefo r e, the prediction of therapeutic peptides is essential for researching on peptide-based therapies. The method of using biological experiments is considered to be time-consuming and labor-intensive. As a fast and accurate method, deep learning can process massive amounts of data on therapeutic peptides. In this research, we raise a deep learning model called Pep-CNN to accurately predict therapeutic peptides. Firstly, we represent the features of the peptide sequence based on the sequence position, the physicochemical property, and the evolutionary-derived feature and use the vectors to represent the sequence. After fusing the features, we use the improved classifier of Convolutional Neural Network (imCNN) to classify and predict eight kinds of peptides. The results show that, compared with other models, Pep-CNN can identify peptides more accurately, which is more conductive to the further research of therapeutic peptides by biomedical scientists. The codes and benchmark datasets are accessible at https://github.com/alivelxj/Pep-CNN. • A new model called Pep-CNN was proposed to predict therapeutic peptides. • The different methods are applied to extract features from the dataset. • An improved convolutional neural network is used to classify the model. [ABSTRACT FROM AUTHOR]

Abstract: Prediction of protein structural class for low-similarity sequences remains a challenging problem. In this study, the new computational method has been developed to predict protein structural class by incorporating alternating word frequency and normalized Lempel–Ziv complexity. To evaluate the performance of the proposed method, jackknife cross-validation tests are performed on three widely used benchmark datasets, 25PDB, 1189 and 640, respectively. We report 83.6%, 81.8% and 83.6% prediction accuracies for 25PDB, 1189 and 640 benchmarks, respectively. Comparison of our results with other methods shows that the proposed method is very promising and may provide a cost-effective alternative to predict protein structural class in particular for low-similarity datasets and may at least play an important complementary role to existing methods. [Copyright &y& Elsevier]

RNA 5-hydroxymethylcytosine (5hmC) modification is the basis of the translation of genetic information and the biological evolution. The study of its distribution in transcriptome is fundamentally crucial to reveal the biological significance of 5hmC. Biochemical experiments can use a variety of sequencing-based technologies to achieve high-throughput identification of 5hmC; however, they are labor-intensive, time-consuming, as well as expensive. Therefore, it is urgent to develop more effective and feasible computational methods. In this paper, a novel and powerful model called iR5hmcSC is designed for identifying 5hmC. Firstly, we extract the different features by K-mer, Pseudo Structure Status Composition and One-Hot encoding. Subsequently, the combination of chi-square test and logistic regression is utilized as the feature selection method to select the optimal feature sets. And then stacking learning, an ensemble learning method including random forest (RF), extra trees (EX), AdaBoost (Ada), gradient boosting decision tree (GBDT), and support vector machine (SVM), is used to recognize 5hmC and non-5hmC. Finally, 10-fold cross-validation test is performed to evaluate the model. The accuracy reaches 85.27% and 79.92% on benchmark dataset and independent dataset, respectively. The result is better than the state-of-the-art methods, which indicates that our model is a feasible tool to identify 5hmC. The datasets and source code are freely available at https://github.com/HongyanShi026/iR5hmcSC. [Display omitted] • A new model named iR5hmcSC was proposed to predict RNA 5hmC sites. • K-mer, Pseudo Structure Status Composition and One-Hot encoding are applied to extract features from the dataset. • A new method combining chi-square test and logistic regression is used to reduce the dimensions of data. • The stacking learning is adopted to classify the model. [ABSTRACT FROM AUTHOR]

Prediction of protein structural classes for low-similarity sequences is useful for understanding fold patterns, regulation, functions, and interactions of proteins. It is well known that feature extraction is significant to prediction of protein structural class and it mainly uses protein primary sequence, predicted secondary structure sequence, and position-specific scoring matrix (PSSM). Currently, prediction solely based on the PSSM has played a key role in improving the prediction accuracy. In this paper, we propose a novel method called CSP-SegPseP-SegACP by fusing consensus sequence (CS), segmented PsePSSM, and segmented autocovariance transformation (ACT) based on PSSM. Three widely used low-similarity datasets (1189, 25PDB, and 640) are adopted in this paper. Then a 700-dimensional (700D) feature vector is constructed and the dimension is decreased to 224D by using principal component analysis (PCA). To verify the performance of our method, rigorous jackknife cross-validation tests are performed on 1189, 25PDB, and 640 datasets. Comparison of our results with the existing PSSM-based methods demonstrates that our method achieves the favorable and competitive performance. This will offer an important complementary to other PSSM-based methods for prediction of protein structural classes for low-similarity sequences. [ABSTRACT FROM AUTHOR]