Cross-Project Defect Prediction (CPDP) utilizes the existing labeled data in the source project to assist with the prediction of unlabeled projects in the target dataset, which effectively improves the prediction performance and has become a research hotspot in software engineering. At present, CPDP can be categorized into homogeneous CPDP and heterogeneous CPDP (HDP), in which HDP doesn’t require that the source project and the target project have the same feature space, thus, it is more widely used in the actual CPDP. Most of current HDP methods map the original features to the latent feature space and reduce the inter-project variation by transferring domain-independent features, but the transferring process ignores the use of domain-related features, which affects the prediction performance of the model. Moreover, the mapped latent features are not conducive to the model’s interpretability. Based on these, this article proposes a Heterogeneous Defect Prediction method based on Feature Disentanglement (FD-HDP). We disentangle the features using domain-related and domain-independent feature extractors, respectively, to improve the interpretability of the model by maximizing the domain adversarial loss during training and guiding the feature extractors to produce accurate domain-related and domain-independent features. The weighted sum of the prediction results from domain-related and domain-independent predictors is used as the final prediction result of the project during the prediction process, which realizes the combination of domain-independent and domain-related features and effectively improves the prediction performance. In this article, we conducted experiments using four publicly available defect datasets to construct heterogeneous scenarios. The results demonstrate that the FD-HDP model shows significant advantages over state-of-the-art methods in six metrics.

RNA-protein interactions (RPIs) play an important role in several fundamental cellular physiological processes, including cell motility, chromosome replication, transcription and translation, and signaling. Predicting RPI can guide the exploration of cellular biological functions, intervening in diseases, and designing drugs. Given this, this study proposes the RPI-gated graph convolutional network (RPI-GGCN) method for predicting RPI based on the gated graph convolutional neural network (GGCN) and co-regularized variational autoencoder (Co-VAE). First, different types of feature information were extracted from RNA and protein sequences by nine feature extraction methods. Second, Co-VAEs are used to eliminate the redundancy of fused features and generate optimal features. Finally, this study introduces gated cyclic units into graph convolutional networks (GCNs) to construct a model for RPI prediction, which efficiently extracts topological information and improves the model's interpretable feature learning and expression capabilities. In the fivefold cross-validation test, the RPI-GGCN method achieved prediction accuracies of 97.27%, 97.32%, 96.54%, 95.76%, and 94.98% on the RPI369, RPI488, RPI1446, RPI1807, and RPI2241 datasets. To test the generalization performance of the model, we used the model trained on RPI369 to predict the independent NPInter v3.0 dataset and achieved excellent performance in all six independent validation sets. By visualizing the RPI network graph based on the prediction results, we aim to provide a new perspective and reference for studying RPI mechanisms and exploring new RPIs. Extensive experimental results demonstrate that RPI-GGCN can provide an efficient, accurate, and stable RPI prediction method.

Precise targeting of transcription factor binding sites (TFBSs) is essential to comprehending transcriptional regulatory processes and investigating cellular function. Although several deep learning algorithms have been created to predict TFBSs, the models' intrinsic mechanisms and prediction results are difficult to explain. There is still room for improvement in prediction performance. We present DeepSTF, a unique deep-learning architecture for predicting TFBSs by integrating DNA sequence and shape profiles. We use the improved transformer encoder structure for the first time in the TFBSs prediction approach. DeepSTF extracts DNA higher-order sequence features using stacked convolutional neural networks (CNNs), whereas rich DNA shape profiles are extracted by combining improved transformer encoder structure and bidirectional long short-term memory (Bi-LSTM), and, finally, the derived higher-order sequence features and representative shape profiles are integrated into the channel dimension to achieve accurate TFBSs prediction. Experiments on 165 ENCODE chromatin immunoprecipitation sequencing (ChIP-seq) datasets show that DeepSTF considerably outperforms several state-of-the-art algorithms in predicting TFBSs, and we explain the usefulness of the transformer encoder structure and the combined strategy using sequence features and shape profiles in capturing multiple dependencies and learning essential features. In addition, this paper examines the significance of DNA shape features predicting TFBSs. The source code of DeepSTF is available at https://github.com/YuBinLab-QUST/DeepSTF/.

Single-cell omics data are growing at an unprecedented rate, whereas effective integration of them remains challenging due to different sequencing methods, quality, and expression pattern of each omics data. In this study, we propose a universal framework for the integration of single-cell multi-omics data based on graph convolutional network (GCN-SC). Among the multiple single-cell data, GCN-SC usually selects one data with the largest number of cells as the reference and the rest as the query dataset. It utilizes mutual nearest neighbor algorithm to identify cell-pairs, which provide connections between cells both within and across the reference and query datasets. A GCN algorithm further takes the mixed graph constructed from these cell-pairs to adjust count matrices from the query datasets. Finally, dimension reduction is performed by using non-negative matrix factorization before visualization. By applying GCN-SC on six datasets, we show that GCN-SC can effectively integrate sequencing data from multiple single-cell sequencing technologies, species or different omics, which outperforms the state-of-the-art methods, including Seurat, LIGER, GLUER and Pamona.

Interactions between DNA and transcription factors (TFs) play an essential role in understanding transcriptional regulation mechanisms and gene expression. Due to the large accumulation of training data and low expense, deep learning methods have shown huge potential in determining the specificity of TFs-DNA interactions. Convolutional network-based and self-attention network-based methods have been proposed for transcription factor binding sites (TFBSs) prediction. Convolutional operations are efficient to extract local features but easy to ignore global information, while self-attention mechanisms are expert in capturing long-distance dependencies but difficult to pay attention to local feature details. To discover comprehensive features for a given sequence as far as possible, we propose a Dual-branch model combining Self-Attention and Convolution, dubbed as DSAC, which fuses local features and global representations in an interactive way. In terms of features, convolution and self-attention contribute to feature extraction collaboratively, enhancing the representation learning. In terms of structure, a lightweight but efficient architecture of network is designed for the prediction, in particular, the dual-branch structure makes the convolution and the self-attention mechanism can be fully utilized to improve the predictive ability of our model. The experiment results on 165 ChIP-seq datasets show that DSAC obviously outperforms other five deep learning based methods and demonstrate that our model can effectively predict TFBSs based on sequence feature alone. The source code of DSAC is available at https://github.com/YuBinLab-QUST/DSAC/.

The prediction of drug-drug interactions (DDIs) is essential for the development and repositioning of new drugs. Meanwhile, they play a vital role in the fields of biopharmaceuticals, disease diagnosis and pharmacological treatment. This article proposes a new method called DBGRU-SE for predicting DDIs. Firstly, FP3 fingerprints, MACCS fingerprints, Pubchem fingerprints and 1D and 2D molecular descriptors are used to extract the feature information of the drugs. Secondly, Group Lasso is used to remove redundant features. Then, SMOTE-ENN is applied to balance the data to obtain the best feature vectors. Finally, the best feature vectors are fed into the classifier combining BiGRU and squeeze-and-excitation (SE) attention mechanisms to predict DDIs. After applying five-fold cross-validation, The ACC values of DBGRU-SE model on the two datasets are 97.51 and 94.98%, and the AUC are 99.60 and 98.85%, respectively. The results showed that DBGRU-SE had good predictive performance for drug-drug interactions.

Advances in single-cell RNA sequencing (scRNA-seq) technologies allow researchers to analyze the genome-wide transcription profile and to solve biological problems at the individual-cell resolution. However, existing clustering methods on scRNA-seq suffer from high dropout rate and curse of dimensionality in the data. Here, we propose a novel pipeline, scBKAP, the cornerstone of which is a single-cell bisecting K-means clustering method based on an autoencoder network and a dimensionality reduction model MPDR. Specially, scBKAP utilizes an autoencoder network to reconstruct gene expression values from scRNA-seq data to alleviate the dropout issue, and the MPDR model composed of the M3Drop feature selection algorithm and the PHATE dimensionality reduction algorithm to reduce the dimensions of reconstructed data. The dimensionality-reduced data are then fed into the bisecting K-means clustering algorithm to identify the clusters of cells. Comprehensive experiments demonstrate scBKAP's superior performance over nine state-of-the-art single-cell clustering methods on 21 public scRNA-seq datasets and simulated datasets. The source codes and datasets are available at https://github.com/YuBinLab-QUST/scBKAP/ and https://doi.org/10.24433/CO.4592131.v1.

MOTIVATION: Multi-label (ML) protein subcellular localization (SCL) is an indispensable way to study protein function. It can locate a certain protein (such as the human transmembrane protein that promotes the invasion of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)) or expression product at a specific location in a cell, which can provide a reference for clinical treatment of diseases such as coronavirus disease 2019 (COVID-19). RESULTS: The article proposes a novel method named ML-locMLFE. First of all, six feature extraction methods are adopted to obtain protein effective information. These methods include pseudo amino acid composition, encoding based on grouped weight, gene ontology, multi-scale continuous and discontinuous, residue probing transformation and evolutionary distance transformation. In the next part, we utilize the ML information latent semantic index method to avoid the interference of redundant information. In the end, ML learning with feature-induced labeling information enrichment is adopted to predict the ML protein SCL. The Gram-positive bacteria dataset is chosen as a training set, while the Gram-negative bacteria dataset, virus dataset, newPlant dataset and SARS-CoV-2 dataset as the test sets. The overall actual accuracy of the first four datasets are 99.23%, 93.82%, 93.24% and 96.72% by the leave-one-out cross validation. It is worth mentioning that the overall actual accuracy prediction result of our predictor on the SARS-CoV-2 dataset is 72.73%. The results indicate that the ML-locMLFE method has obvious advantages in predicting the SCL of ML protein, which provides new ideas for further research on the SCL of ML protein. AVAILABILITY AND IMPLEMENTATION: The source codes and datasets are publicly available at https://github.com/QUST-AIBBDRC/ML-locMLFE/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Predicting binding affinities of peptide antigens presented on major histocompatibility complex (MHC) is of great importance in T-cell immune response research. Accurate prediction of peptide-MHC binding affinities is essential for vaccine design and disease treatment. Recent deep learning-based prediction methods have shown that effective sequence embedding is critical to accurately predict binding affinities. One common neural network layer shared by these methods is the global average pooling layer that aggregates features. However, can we design a better global pooling layer? Here, we introduce a novel cross attention pooling (caPool) layer to aggregate features. As our initial application of caPool, a novel end-to-end transformer model, called capTransformer, is proposed for peptide-MHC class I binding prediction. In our model, caPool jointly aligns peptide-MHC residual pairs and aggregates residual features. Thus, instead of treating all residues equally and independently, caPool focuses more on correlated residue pairs that are potentially contact pairs contributing major forces to stabilize the complex structure. Using a five-fold cross-validation experiment, we found that caPool achieved the highest PCC value of 0.845, which was 0.139 higher than a global average pooling. Here, the global pooling layer was the only difference between the two tested models, and this observation indicated that global average pooling was not always the best choice. Importantly, our capTransformer model achieved a SRCC value of 0.614 (i.e., 6.4% higher than the best-performing method) when applied to the IEDB dataset.

Bacterial genomes are now recognized as interacting intimately with cellular processes. Uncovering organizational mechanisms of bacterial genomes has been a primary focus of researchers to reveal the potential cellular activities. The advances in both experimental techniques and computational models provide a tremendous opportunity for understanding these mechanisms, and various studies have been proposed to explore the organization rules of bacterial genomes associated with functions recently. This review focuses mainly on the principles that shape the organization of bacterial genomes, both locally and globally. We first illustrate local structures as operons/transcription units for facilitating co-transcription and horizontal transfer of genes. We then clarify the constraints that globally shape bacterial genomes, such as metabolism, transcription and replication. Finally, we highlight challenges and opportunities to advance bacterial genomic studies and provide application perspectives of genome organization, including pathway hole assignment and genome assembly and understanding disease mechanisms.

The rapid development of single-cell RNA sequencing (scRNA-Seq) technology provides strong technical support for accurate and efficient analyzing single-cell gene expression data. However, the analysis of scRNA-Seq is accompanied by many obstacles, including dropout events and the curse of dimensionality. Here, we propose the scGMAI, which is a new single-cell Gaussian mixture clustering method based on autoencoder networks and the fast independent component analysis (FastICA). Specifically, scGMAI utilizes autoencoder networks to reconstruct gene expression values from scRNA-Seq data and FastICA is used to reduce the dimensions of reconstructed data. The integration of these computational techniques in scGMAI leads to outperforming results compared to existing tools, including Seurat, in clustering cells from 17 public scRNA-Seq datasets. In summary, scGMAI is an effective tool for accurately clustering and identifying cell types from scRNA-Seq data and shows the great potential of its applicative power in scRNA-Seq data analysis. The source code is available at https://github.com/QUST-AIBBDRC/scGMAI/.

MOTIVATION: Mitochondria are an essential organelle in most eukaryotes. They not only play an important role in energy metabolism but also take part in many critical cytopathological processes. Abnormal mitochondria can trigger a series of human diseases, such as Parkinson's disease, multifactor disorder and Type-II diabetes. Protein submitochondrial localization enables the understanding of protein function in studying disease pathogenesis and drug design. RESULTS: We proposed a new method, SubMito-XGBoost, for protein submitochondrial localization prediction. Three steps are included: (i) the g-gap dipeptide composition (g-gap DC), pseudo-amino acid composition (PseAAC), auto-correlation function (ACF) and Bi-gram position-specific scoring matrix (Bi-gram PSSM) are employed to extract protein sequence features, (ii) Synthetic Minority Oversampling Technique (SMOTE) is used to balance samples, and the ReliefF algorithm is applied for feature selection and (iii) the obtained feature vectors are fed into XGBoost to predict protein submitochondrial locations. SubMito-XGBoost has obtained satisfactory prediction results by the leave-one-out-cross-validation (LOOCV) compared with existing methods. The prediction accuracies of the SubMito-XGBoost method on the two training datasets M317 and M983 were 97.7% and 98.9%, which are 2.8-12.5% and 3.8-9.9% higher than other methods, respectively. The prediction accuracy of the independent test set M495 was 94.8%, which is significantly better than the existing studies. The proposed method also achieves satisfactory predictive performance on plant and non-plant protein submitochondrial datasets. SubMito-XGBoost also plays an important role in new drug design for the treatment of related diseases. AVAILABILITY AND IMPLEMENTATION: The source codes and data are publicly available at https://github.com/QUST-AIBBDRC/SubMito-XGBoost/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: The prediction of protein-protein interaction (PPI) sites is a key to mutation design, catalytic reaction and the reconstruction of PPI networks. It is a challenging task considering the significant abundant sequences and the imbalance issue in samples. RESULTS: A new ensemble learning-based method, Ensemble Learning of synthetic minority oversampling technique (SMOTE) for Unbalancing samples and RF algorithm (EL-SMURF), was proposed for PPI sites prediction in this study. The sequence profile feature and the residue evolution rates were combined for feature extraction of neighboring residues using a sliding window, and the SMOTE was applied to oversample interface residues in the feature space for the imbalance problem. The Multi-dimensional Scaling feature selection method was implemented to reduce feature redundancy and subset selection. Finally, the Random Forest classifiers were applied to build the ensemble learning model, and the optimal feature vectors were inserted into EL-SMURF to predict PPI sites. The performance validation of EL-SMURF on two independent validation datasets showed 77.1% and 77.7% accuracy, which were 6.2-15.7% and 6.1-18.9% higher than the other existing tools, respectively. AVAILABILITY AND IMPLEMENTATION: The source codes and data used in this study are publicly available at http://github.com/QUST-AIBBDRC/EL-SMURF/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Using knowledge-based interpretation to analyze omics data can not only obtain essential information regarding various biological processes, but also reflect the current physiological status of cells and tissue. The major challenge to analyze gene expression data, with a large number of genes and small samples, is to extract disease-related information from a massive amount of redundant data and noise. Gene selection, eliminating redundant and irrelevant genes, has been a key step to address this problem. RESULTS: The modified method was tested on four benchmark datasets with either two-class phenotypes or multiclass phenotypes, outperforming previous methods, with relatively higher accuracy, true positive rate, false positive rate and reduced runtime. CONCLUSIONS: This paper proposes an effective feature selection method, combining double RBF-kernels with weighted analysis, to extract feature genes from gene expression data, by exploring its nonlinear mapping ability.

The task of face recognition has been actively researched in recent years because of its many applications in various domains. This paper presents a robust face recognition system using curvelet-based two-dimensional principle component analysis (2D PCA) to address the problem of human face recognition from still images. 2D PCA has advantages over PCA in evaluating the covariance matrix accurately and time complexity. Inspired by the attractive attributes of curvelets in catching the edge singularities with very few coefficients in a non-adaptive manner, we introduce the scheme of decomposing images into curvelet subbands and applying 2D PCA to create a representative feature set. Experiments were designed with different implementations of each module using standard testing database. We experimented with changing the illumination normalization procedure; comparing the baseline PCA-based method with the proposed scheme; studying effects on algorithm performance of k-nearest neighbor (kNN) classifier and Support Vector Machine (SVM) classifier in the classification process; also we experimented with different databases such as FERET, etc. High accuracy rate were achieved by the proposed scheme through a comparative study.

Document image segmentation is an important research area of document image analysis which classifies the contents of a document image into a set of text and non-text classes. Previous existing methods are often designed to classify text and halftone therefore they perform poorly in classifying graphics, tables and circuit, etc. In this paper, we present a robust multi-level classification method using multi-layer perceptron (MLP) and support vector machine (SVM) to segment the texts from non-texts and thereafter classify them as tables, graphics and halftones. This method outperforms previously existing methods by overcoming various issues associated with the complexity of document images. Experimental results prove the effectiveness of our proposed method. By virtue of our multi-level classification approach, the text components, halftone components, graphic components and table components are accurately classified respectively which would highly improve OCR accuracy to reduce garbage symbols as well as increase compression ratio thereafter simultaneously.