Protein toxicity prediction is crucial for drug discovery, safety assessment, and toxicological research. This study introduces $\mathrm{Deep\_{T}PPred}$, a novel hybrid deep learning (DL) model that integrates Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) for accurate protein toxicity prediction. The model effectively combines diverse protein sequence descriptors to capture complex sequence relationships by leveraging a feature fusion technique. The methodology involved advanced feature extraction, rigorous training, and performance evaluation using benchmark datasets. $\mathrm{Deep\_{T}PPred}$ demonstrates state-of-the-art performance with an accuracy of 0.9983, specificity of 0.9988, sensitivity of 0.9975, and Kappa and MCC values of 0.9963. These results underscore the proposed model's robustness, reliability, and generalization capability, surpassing existing models across all metrics. The study highlights the potential of hybrid DL and feature fusion techniques to significantly enhance protein toxicity prediction, providing valuable insights and tools for bioinformatics pipelines and applications.

Abstract Breast cancer is one of the leading causes of cancer-related morbidity worldwide, underscoring the need for advanced diagnostic tools to improve early detection and treatment outcomes. This study introduces MammoSegNet, a novel convolutional neural network architecture optimized for precisely segmenting mammographic images. The proposed MammoSegNet incorporates Inception-ResNet blocks, Squeeze-and-Excitation (SE) modules, and dilated convolutions to enable multi-scale feature extraction and efficient attention refinement while maintaining low computational complexity. MammoSegNet performance was rigorously evaluated on BCDR-D01 and INbreast datasets to examine its robustness and generalization. Using stratified fivefold cross-validation, the model was trained on BCDR-D01 and tested on the unseen INbreast dataset through Monte Carlo cross-validation. Preprocessing techniques, including Region of Interest (ROI) Isolation to concentrate on relevant areas, Normalization to standardized pixel intensities, and Data Augmentation to expand the dataset and enhance the model’s robustness, were employed. Additionally, a specialized image enhancement method called peak feature intensity transformation (PFIT) was designed to amplify diagnostic features while preserving structural integrity. Comparative evaluations confirmed MammoSegNet’s superior performance across metrics, achieving 97% accuracy on BCDR-D01 and 95% on INbreast. Statistical t-tests validated these improvements, and visual heatmaps demonstrated the model’s effectiveness in isolating tumor regions. These findings establish MammoSegNet as a promising tool for enhancing breast cancer diagnostic accuracy and reliability in medical applications.

Druggable proteins are defined as proteins that can interact with drugs to modulate certain biological activity. The identification of druggable proteins holds significant clinical importance, directly impacting the development of targeted therapies for diseases like cancer and metabolic disorders. Identifying druggable proteins involves various methods, including computational prediction models, mass spectrometry (MS), and biochemical assays, but achieving high accuracy remains a challenge. This study proposes DrugEL, an ensemble learning model that uses Bayesian inference to integrate predictions from multiple algorithms: Random Forest (RF), K-Nearest Neighbor (KNN), LightGBM (LGBM), and Decision Tree (DT) with seven feature extraction methods (LSA, AAC, PAAC, GAAC, NMBroto, AAIndex, and KNN). The results show that DrugEL outperforms existing models in terms of accuracy (0.9758), MCC (0.9515), AUC (0.9758), sensitivity (0.9742) and specificity (0.9774), particularly excelling with the LSA method.

Security concerns for Internet of Things (IoT) applications have been alarming because of their widespread use in different enterprise systems. The potential threats to these applications are constantly emerging and changing, and, therefore, sophisticated and dependable defense solutions are necessary against such threats. With the rapid development of IoT networks and evolving threat types, the traditional machine learning based IDS must update to cope with the security requirements of the current sustainable IoT environment. In recent years, deep learning and deep transfer learning have progressed and experienced great success in different fields and have emerged as a potential solution for dependable network intrusion detection. However, new and emerging challenges have arisen related to the accuracy, efficiency, scalability, and dependability of the traditional IDS in a heterogeneous IoT setup. This manuscript proposes a deep transfer learning based dependable IDS model that outperforms several existing approaches. The unique contributions include effective attribute selection, which is best suited to identify normal and attack scenarios for a small amount of labeled data, designing a dependable deep transfer learning based ResNet model and evaluating considering real-world data. To this end, a comprehensive experimental performance evaluation has been conducted. Extensive analysis and performance evaluation show that the proposed model is robust, more efficient, and has demonstrated better performance, ensuring dependability.

Human skin disease, the most infectious dermatological ailment globally, is initially diagnosed by sight. Some clinical screening and dermoscopic analysis of skin biopsies and scrapings for accurate classification are medically compulsory. Classification of skin diseases using medical images is more challenging because of the complex formation and variant colors of the disease and data security concerns. Both the Convolution Neural Network (CNN) for classification and a federated learning approach for data privacy preservation show significant performance in the realm of medical imaging fields. In this paper, a custom image dataset was prepared with four classes of skin disease, a CNN model was suggested and compared with several benchmark CNN algorithms, and an experiment was carried out to ensure data privacy using a federated learning approach. An image augmentation strategy was followed to enlarge the dataset and make the model more general. The proposed model achieved a precision of 86%, 43%, and 60%, and a recall of 67%, 60%, and 60% for acne, eczema, and psoriasis. In the federated learning approach, after distributing the dataset among 1000, 1500, 2000, and 2500 clients, the model showed an average accuracy of 81.21%, 86.57%, 91.15%, and 94.15%. The CNN-based skin disease classification merged with the federated learning approach is a breathtaking concept to classify human skin diseases while ensuring data security.

Erectile dysfunction (ED) can be caused by different diseases and controlled by several genetic networks. In this study, to identify the genes related to ED, the expression profiles of normal and ED samples were investigated by the Gene Expression Omnibus (GEO) database. Seventeen genes were identified as associated genes with ED. The protein and nucleic acid sequences of selected genes were retrieved from the UCSC database. Selected genes were diverse according to their physicochemical properties and functions. Category function revealed that selected genes are involved in pathways related to humans some diseases. Furthermore, based on protein interactions, genes associated with the insulin pathway had the greatest interaction with the studied genes. To identify the common cis-regulatory elements, the promoter site of the selected genes was retrieved from the UCSC database. The Gapped Local Alignment of Motifs tool was used for finding common conserved motifs into the promoter site of selected genes. Besides, INSR protein as an insulin receptor precursor showed a high potential site for posttranslation modifications, including phosphorylation and N-glycosylation. Also, in this study, two Guanine-Cytosine (GC)-rich regions were identified as conserved motifs in the upstream of studied genes which can be involved in regulating the expression of genes associated with ED. Also, the conserved binding site of miR-29-3p that is involved in various cancers was observed in the 3' untranslated region of genes associated with ED. Our study introduced new genes associated with ED, which can be good candidates for further analyzing related to human ED.

Discovering drug-target (protein) interactions (DTIs) is of great significance for researching and developing novel drugs, having a tremendous advantage to pharmaceutical industries and patients. However, the prediction of DTIs using wet-lab experimental methods is generally expensive and time-consuming. Therefore, different machine learning-based methods have been developed for this purpose, but there are still substantial unknown interactions needed to discover. Furthermore, data imbalance and feature dimensionality problems are a critical challenge in drug-target datasets, which can decrease the classifier performances that have not been significantly addressed yet. This paper proposed a novel drug-target interaction prediction method called PreDTIs. First, the feature vectors of the protein sequence are extracted by the pseudo-position-specific scoring matrix (PsePSSM), dipeptide composition (DC) and pseudo amino acid composition (PseAAC); and the drug is encoded with MACCS substructure fingerings. Besides, we propose a FastUS algorithm to handle the class imbalance problem and also develop a MoIFS algorithm to remove the irrelevant and redundant features for getting the best optimal features. Finally, balanced and optimal features are provided to the LightGBM Classifier to identify DTIs, and the 5-fold CV validation test method was applied to evaluate the prediction ability of the proposed method. Prediction results indicate that the proposed model PreDTIs is significantly superior to other existing methods in predicting DTIs, and our model could be used to discover new drugs for unknown disorders or infections, such as for the coronavirus disease 2019 using existing drugs compounds and severe acute respiratory syndrome coronavirus 2 protein sequences.

This study aimed to identify significant gene expression profiles of the human lung epithelial cells caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections. We performed a comparative genomic analysis to show genomic observations between SARS-CoV and SARS-CoV-2. A phylogenetic tree has been carried for genomic analysis that confirmed the genomic variance between SARS-CoV and SARS-CoV-2. Transcriptomic analyses have been performed for SARS-CoV-2 infection responses and pulmonary arterial hypertension (PAH) patients' lungs as a number of patients have been identified who faced PAH after being diagnosed with coronavirus disease 2019 (COVID-19). Gene expression profiling showed significant expression levels for SARS-CoV-2 infection responses to human lung epithelial cells and PAH lungs as well. Differentially expressed genes identification and integration showed concordant genes (SAA2, S100A9, S100A8, SAA1, S100A12 and EDN1) for both SARS-CoV-2 and PAH samples, including S100A9 and S100A8 genes that showed significant interaction in the protein-protein interactions network. Extensive analyses of gene ontology and signaling pathways identification provided evidence of inflammatory responses regarding SARS-CoV-2 infections. The altered signaling and ontology pathways that have emerged from this research may influence the development of effective drugs, especially for the people with preexisting conditions. Identification of regulatory biomolecules revealed the presence of active promoter gene of SARS-CoV-2 in Transferrin-micro Ribonucleic acid (TF-miRNA) co-regulatory network. Predictive drug analyses provided concordant drug compounds that are associated with SARS-CoV-2 infection responses and PAH lung samples, and these compounds showed significant immune response against the RNA viruses like SARS-CoV-2, which is beneficial in therapeutic development in the COVID-19 pandemic.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is accountable for the cause of coronavirus disease (COVID-19) that causes a major threat to humanity. As the spread of the virus is probably getting out of control on every day, the epidemic is now crossing the most dreadful phase. Idiopathic pulmonary fibrosis (IPF) is a risk factor for COVID-19 as patients with long-term lung injuries are more likely to suffer in the severity of the infection. Transcriptomic analyses of SARS-CoV-2 infection and IPF patients in lung epithelium cell datasets were selected to identify the synergistic effect of SARS-CoV-2 to IPF patients. Common genes were identified to find shared pathways and drug targets for IPF patients with COVID-19 infections. Using several enterprising Bioinformatics tools, protein-protein interactions (PPIs) network was designed. Hub genes and essential modules were detected based on the PPIs network. TF-genes and miRNA interaction with common differentially expressed genes and the activity of TFs are also identified. Functional analysis was performed using gene ontology terms and Kyoto Encyclopedia of Genes and Genomes pathway and found some shared associations that may cause the increased mortality of IPF patients for the SARS-CoV-2 infections. Drug molecules for the IPF were also suggested for the SARS-CoV-2 infections.