MOTIVATION: Named entity recognition (NER) is a fundamental component of structured knowledge extraction, yet its effectiveness in emerging domains remains by the scarcity of high-quality, domain-specific annotated corpora. Although data augmentation and distant supervision have been explored to alleviate this issue, existing methods often introduce limited entity diversity, noisy labels, or disrupt contextual integrity, thereby limiting their generalization ability in low-resource settings. RESULTS: In this study, we propose DA-BioNER, a context-preserving data expansion framework for biomedical NER. DA-BioNER combines multiple base NER models trained on few-shot data to provide coarse annotations, followed by refinement using a large language model (LLM) guided by global biomedical knowledge. Unlike generation-based augmentation methods that synthesize new sentences, DA-BioNER performs annotation refinement within existing sentences, preserving both syntactic structure and semantic context. By constraining the role of LLM to refinement rather than open-ended generation, the framework effectively reduces hallucination while improving label precision and consistency. We evaluate DA-BioNER on three benchmark datasets (NCBI-Disease, BC5CDR, and BioRED), under low-resource conditions. In 40-shot settings, DA-BioNER achieves F1-scores of 0.750, 0.795, and 0.799, respectively, outperforming state-of-the-art methods, including LSMS, DAGA, and MELM, by up to 0.32. Under more extreme few-shot settings, DA-BioNER further improves F1-scores by up to 0.08, while generating an average of 1,391 additional unique entities, substantially enriching training diversity. These results demonstrate that DA-BioNER provides a scalable and adaptable solution for robust biomedical NER, particularly in domain adaptation and low-resource scenarios. AVAILABILITY: DA-BioNER is publicly available at https://github.com/DMnBI/DA-BioNER.

MOTIVATION: Mobile genetic elements (MGEs) play an important role in facilitating the acquisition of antibiotic resistance genes (ARGs) within microbial communities, significantly impacting the evolution of antibiotic resistance. Understanding the mechanism and trajectory of ARG acquisition requires a comprehensive analysis of the ARG-carrying mobilome-a collective set of MGEs carrying ARGs. However, identifying the mobilome within complex microbiomes poses considerable challenges. Existing MGE prediction methods, designed primarily for single genomes, exhibit substantial limitations when applied to metagenomic data, often producing high false positive rates in identifying target MGEs from metagenome sequencing data. RESULTS: To address these challenges, we developed DeepMobilome, a novel approach for accurately identifying target MGEs within the microbiome. DeepMobilome leverages a convolutional neural network trained on read alignment data derived from sequence alignment map (SAM) files, providing superior accuracy in detecting MGEs. Trained on 364 647 cases, DeepMobilome achieved a high validation accuracy of 0.99. DeepMobilome consistently outperformed existing methods in discerning the presence of target MGE sequences across diverse test sets. In single-genome test scenarios, DeepMobilome showed an F1-score of 0.935, compared to 0.755 and 0.670 for MGEfinder and ISMapper, respectively, demonstrating its substantial improvements in prediction accuracy. Extensive evaluations across simulated microbiomes further validated the robustness and reliability of DeepMobilome in practical applications. In real microbiome data, DeepMobilome successfully identified six ARG-carrying MGEs across diverse populations. By addressing the limitations of current methods, DeepMobilome offers a powerful tool for advancing our understanding of ARG dissemination and supports targeted interventions in combating antibiotic resistance.

Viruses are ubiquitous in humans and various environments and continually mutate themselves. Identifying viruses in an environment without cultivation is challenging; however, promoting the screening of novel viruses and expanding the knowledge of viral space is essential. Homology-based methods that identify viruses using known viral genomes rely on sequence alignments, making it difficult to capture remote homologs of the known viruses. To accurately capture viral signals from metagenomic samples, models are needed to understand the patterns encoded in the viral genomes. In this study, we developed a hierarchical BERT model named ViBE to detect eukaryotic viruses from metagenome sequencing data and classify them at the order level. We pre-trained ViBE using read-like sequences generated from the virus reference genomes and derived three fine-tuned models that classify paired-end reads to orders for eukaryotic deoxyribonucleic acid viruses and eukaryotic ribonucleic acid viruses. ViBE achieved higher recall than state-of-the-art alignment-based methods while maintaining comparable precision. ViBE outperformed state-of-the-art alignment-free methods for all test cases. The performance of ViBE was also verified using real sequencing datasets, including the vaginal virome.

BACKGROUND: Diverse microbiome communities drive biogeochemical processes and evolution of animals in their ecosystems. Many microbiome projects have demonstrated the power of using metagenomics to understand the structures and factors influencing the function of the microbiomes in their environments. In order to characterize the effects from microbiome composition for human health, diseases, and even ecosystems, one must first understand the relationship of microbes and their environment in different samples. Running machine learning model with metagenomic sequencing data is encouraged for this purpose, but it is not an easy task to make an appropriate machine learning model for all diverse metagenomic datasets. RESULTS: We introduce MegaR, an R Shiny package and web application, to build an unbiased machine learning model effortlessly with interactive visual analysis. The MegaR employs taxonomic profiles from either whole metagenome sequencing or 16S rRNA sequencing data to develop machine learning models and classify the samples into two or more categories. It provides various options for model fine tuning throughout the analysis pipeline such as data processing, multiple machine learning techniques, model validation, and unknown sample prediction that can be used to achieve the highest prediction accuracy possible for any given dataset while still maintaining a user-friendly experience. CONCLUSIONS: Metagenomic sample classification and phenotype prediction is important particularly when it applies to a diagnostic method for identifying and predicting microbe-related human diseases. MegaR provides various interactive visualizations for user to build an accurate machine-learning model without difficulty. Unknown sample prediction with a properly trained model using MegaR will enhance researchers to identify the sample property in a fast turnaround time.

With the widespread use of smartphones, the number of malware has been increasing exponentially. Among smart devices, android devices are the most targeted devices by malware because of their high popularity. This paper proposes a novel framework for android malware detection. Our framework uses various kinds of features to reflect the properties of android applications from various aspects, and the features are refined using our existence-based or similarity-based feature extraction method for effective feature representation on malware detection. Besides, a multimodal deep learning method is proposed to be used as a malware detection model. This paper is the first study of the multimodal deep learning to be used in the android malware detection. With our detection model, it was possible to maximize the benefits of encompassing multiple feature types. To evaluate the performance, we carried out various experiments with a total of 41 260 samples. We compared the accuracy of our model with that of other deep neural network models. Furthermore, we evaluated our framework in various aspects including the efficiency in model updates, the usefulness of diverse features, and our feature representation method. In addition, we compared the performance of our framework with those of other existing methods including deep learning-based methods.

Smart TVs are rapidly replacing conventional TVs. In a number of countries, set-top boxes (STB) are widely used to relay TV channels to smart TVs. In such cases, smart TVs cannot identify which TV channel they are receiving. This situation makes it challenging for smart TVs to provide their users with a variety of personalized services, such as context-aware services and recommendation services. In this paper, we introduce our TV channel matching system that resolves such problems. We propose strategies for scaling-out the matching system and improving its accuracy.

UNLABELLED: : MGEScan-long terminal repeat (LTR) and MGEScan-non-LTR are successfully used programs for identifying LTRs and non-LTR retrotransposons in eukaryotic genome sequences. However, these programs are not supported by easy-to-use interfaces nor well suited for data visualization in general data formats. Here, we present MGEScan, a user-friendly system that combines these two programs with a Galaxy workflow system accelerated with MPI and Python threading on compute clusters. MGEScan and Galaxy empower researchers to identify transposable elements in a graphical user interface with ready-to-use workflows. MGEScan also visualizes the custom annotation tracks for mobile genetic elements in public genome browsers. A maximum speed-up of 3.26× is attained for execution time using concurrent processing and MPI on four virtual cores. MGEScan provides four operational modes: as a command line tool, as a Galaxy Toolshed, on a Galaxy-based web server, and on a virtual cluster on the Amazon cloud. AVAILABILITY AND IMPLEMENTATION: MGEScan tutorials and source code are available at http://mgescan.readthedocs.org/ CONTACT: [email protected] or [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: One of the difficulties in metagenomic assembly is that homologous genes from evolutionarily closely related species may behave like repeats and confuse assemblers. As a result, small contigs, each representing a short gene fragment, instead of complete genes, may be reported by an assembler. This further complicates annotation of metagenomic datasets, as annotation tools (such as gene predictors or similarity search tools) typically perform poorly on configs encoding short gene fragments. RESULTS: We present a novel way of using the de Bruijn graph assembly of metagenomes to improve the assembly of genes. A network matching algorithm is proposed for matching the de Bruijn graph of contigs against reference genes, to derive 'gene paths' in the graph (sequences of contigs containing gene fragments) that have the highest similarities to known genes, allowing gene fragments contained in multiple contigs to be connected to form more complete (or intact) genes. Tests on simulated and real datasets show that our approach (called GeneStitch) is able to significantly improve the assembly of genes from metagenomic sequences, by connecting contigs with the guidance of homologous genes-information that is orthogonal to the sequencing reads. We note that the improvement of gene assembly can be observed even when only distantly related genes are available as the reference. We further propose to use 'gene graphs' to represent the assembly of reads from homologous genes and discuss potential applications of gene graphs to improving functional annotation for metagenomics. AVAILABILITY: The tools are available as open source for download at http://omics.informatics.indiana.edu/GeneStitch CONTACT: [email protected].

BACKGROUND: Modern pyrosequencing techniques make it possible to study complex bacterial populations, such as 16S rRNA, directly from environmental or clinical samples without the need for laboratory purification. Alignment of sequences across the resultant large data sets (100,000+ sequences) is of particular interest for the purpose of identifying potential gene clusters and families, but such analysis represents a daunting computational task. The aim of this work is the development of an efficient pipeline for the clustering of large sequence read sets. METHODS: Pairwise alignment techniques are used here to calculate genetic distances between sequence pairs. These methods are pleasingly parallel and have been shown to more accurately reflect accurate genetic distances in highly variable regions of rRNA genes than do traditional multiple sequence alignment (MSA) approaches. By utilizing Needleman-Wunsch (NW) pairwise alignment in conjunction with novel implementations of interpolative multidimensional scaling (MDS), we have developed an effective method for visualizing massive biosequence data sets and quickly identifying potential gene clusters. RESULTS: This study demonstrates the use of interpolative MDS to obtain clustering results that are qualitatively similar to those obtained through full MDS, but with substantial cost savings. In particular, the wall clock time required to cluster a set of 100,000 sequences has been reduced from seven hours to less than one hour through the use of interpolative MDS. CONCLUSIONS: Although work remains to be done in selecting the optimal training set size for interpolative MDS, substantial computational cost savings will allow us to cluster much larger sequence sets in the future.