National University of Computer and Emerging Sciences

There is a great need to categorize cancer types for early cancer detection and treatment. RNA-Seq data is essential for getting insight into the differentially expressed genes. Due to its high dimensionality and complexity, performing an analysis on RNA-Seq data is quite challenging. In the past, RNA-Seq data were analyzed for a single cancer type as a two-class problem (either positive or negative) and did not contain information from other classes of cancer types. To classify different cancer types and discover the most promising genes, RNA-Seq data for different types of cancer should be examined. Multiple repositories offer RNA-Seq-based cancer types data. The present study incorporates a dataset from the Mendeley repository for classification. RNA-Seq values are then converted to their respective 2D images using some transformations. The classification problem is handled by five Transfer Learning (TL) algorithms (VGG16, VGG19, Resnet50, Resnet101, and Resnet152). Four different splitting strategies are applied for each classifier presented in the results and discussion section. A comparative analysis is also carried out with and without data augmentation. Results show that classifiers perform best at a split of 70-30. VGG16 attained the best position on overall results by achieving an accuracy of 95%. Hence, VGG16 is the leading TL algorithm for classification among all the accessible models and is not difficult to execute and easy to comprehend.

This study produces a novel hybrid model to predict traffic flow breakdowns. The procedure started with the creation of a Percentile Density Approach (PDA) which identified traffic flow breakdown using a percentile -based analysis of density. In this method, traffic flow breakdown was found between traffic free-flow and congested states. Afterwards, a fuzzy logic approach was developed to advance the process into a model form whilst incorporating other factors. As a result, this model integrates the major parameters in the PDA, such as free-flow and congested density as well as weather conditions, driver behavior, and seasonal factors. Thus, the fuzzy hybrid logic system developed, uses membership functions built around both the PDA and these additional parameters enhanced by well-defined sets of crisps decision rules. This resulted in a model which describes traffic flow breakdown incorporating fundamental traffic variables with real-world uncertainties

Abstract A definitive correlate of protection (CoP) for SARS-CoV-2 has yet to be formally established. Previously, using data from a series of non-human primate vaccine challenge studies, we reported that neutralising antibodies (NAbs) are the strongest candidate for clinical protection against COVID-19 and that spike binding antibody is the strongest candidate CoP for viral burden post-challenge. In this study, we further characterised the protective binding antibody profile by analysing spike antibody-dependent complement deposition, FcγR binding, isotype and antibody glycosylation. Using the machine learning platform SIMON, we demonstrate that antibody-dependent complement deposition (ADCD) and FcγR binding are strong candidate co-correlates for each of the post-challenge outcomes; viral load and lung pathology. We found that spike antibody sialylation closely followed by FcγR2A, was the spike antibody feature with the strongest negative correlation with histopathology score. Spike antibody ADCD, FcγR binding, isotype and glycosylation significantly differed by immunisation regimen and sex, which demonstrates the heterogeneity of immune mechanisms induced by different immunisation platforms. We conclude that spike binding antibody, with the protective functional characteristics described herein, is a candidate CoP that captures both protection from severe clinical disease and protection against a high viral burden. These findings should be taken into consideration for future SARS-CoV-2 vaccine development.

Abstract The sustainability of Atlantic salmon ( Salmo salar ) farming is threatened by infectious diseases, environmental stressors, feed limitations, and regulatory or economic constraints. Although current health monitoring is improving with AI-powered camera systems that analyze behavior and nutrition, these tools typically identify stress responses rather than early signs of disease states. Because the microbial community undergoes successional reassembly in response to physiological disruptions before host barriers are breached, the microbiome offers a proactive early warning approach. In this study, a comparative cohort design was employed using a total of 171 individuals (n= 85 “healthy”; n=86 “lesioned”) collected from a commercial marine facility and classified based on external clinical signs. The microbiome of multiple body sites (gills, skin, urogenital pore, and mucosa) from healthy and lesioned salmon were profiled using 16S rRNA amplicon sequencing. No differences in alpha diversity were observed between tissues and conditions. However, beta diversity was significantly different in clinical status, and the interaction of tissue with the status. Conversely, the mucosa and urogenital microbiomes were compositionally similar to each other, as were the gill and skin microbiomes, suggesting that urogenital swabs could serve as a non-invasive proxy for gut microbiome profiling, and skin for gill microbiomes. Several supervised models were trained on these profiles and used to classify salmon status with high accuracy. Based on these data, two Salmon Microbiome Health Score were developed that accurately differentiated the two cohorts. These scores are proposed as a novel biomarker, enabling proactive health management in aquaculture and complementing emerging technological monitoring systems.

The growth of Internet of Things (IoT) technologies, such as cloud computing, 5G communication, and wireless sensor networks, is driving a smarter and more connected future. Thousands of terabytes of data are uploaded to cloud servers each day for storage or computation. Due to data privacy, we cannot upload personal pictures, videos, locations, and medical records directly to the cloud because they will be at risk if compromised. Due to the untrusted nature of the cloud, data needs to be encrypted to ensure confidentiality before being outsourced to it. The data must first be decrypted before any operation can be performed, which can be resource-intensive and wasteful. Secure data transmission from sensors to an Internet host becomes a critical issue for the success of IoT. To address these issues, this paper introduces a lightweight certificateless signcryption scheme with an equality test (CLS-ET), which leverages the power of hyperelliptic curves. This scheme obtains the security goals of authentication, integrity, confidentiality, and non-repudiation in one logical step. Furthermore, this scheme enables us to verify whether two ciphertexts are encrypted with the same or different keys that contain the same information without decrypting them. Indistinguishability under adaptive chosen ciphertext attack (IND-CCA2), existential unforgeability under chosen message attack (EUF-CMA), and one-wayness under adaptive chosen ciphertext attack (OW-CCA2) level security have been achieved by the proposed scheme in the Random Oracle Model (ROM). Furthermore, we compared our proposed scheme with other existing state-of-the-art schemes. While maintaining security and functionality, our scheme reduces computation costs for encryption, decryption, and testing stages, thereby improving efficiency in resource-constrained IoT-enabled Wireless Body Area Networks. • A lightweight certificateless signcryption scheme is proposed for WBAN security. • An equality test enables comparison of ciphertexts without revealing plaintext. • The scheme uses hyperelliptic curve cryptography for computational efficiency. • Achieves IND-CCA2 and EUF-CMA security in the Random Oracle Model. • Reduces overhead in encryption, decryption, and equality testing phases.

Abstract In developing countries, agriculture plays an essential role in economic growth by providing food security, employment, and raw materials for industries. Diseases in plants, as in other agricultural groups, have a great impact on the reduction of global crop yields. This indicates a need for modern precision agriculture to perform their identification fast and accurately. Nowadays, Artificial Intelligence (AI) and deep-learning algorithms are used to detect diseases from leaves images. These approaches are significantly better than traditional methods. Nevertheless, visually similar diseases identification and high accuracy gains involving plenty of categories continue to be challenging. In this research, we used plant village dataset and applied an efficient and enhanced deep learning approach to classify 38 distinct plant leaf diseases. EfficientNetB3 based Convolutional Neural Network (ENBCNN) is selected as a generalizing feature extractor and produces a distinctive classification layer projecting fine patterns of disease with high discriminability to support differentiation. The results of the experiments showed great accuracy 99.7%, very stable learning curves, and reliable cross-validation effects. Our proposed method can accurately identify a wide spectrum of diseases. Therefore, developing this system, in practical way to implement it into mobile devices and increase the data to improve its usefulness are real conditions.