Search for: All records

Masked Autoencoder (MAE) is a notable method for self-supervised pretraining in visual representation learning. It operates by randomly masking image patches and reconstructing these masked patches using the unmasked ones. A key limitation of MAE lies in its disregard for the varying informativeness of different patches, as it uniformly selects patches to mask. To overcome this, some approaches propose masking based on patch informativeness. However, these methods often do not consider the specific requirements of downstream tasks, potentially leading to suboptimal representations for these tasks. In response, we introduce the Multi-level Optimized Mask Autoencoder (MLO-MAE), a novel framework that leverages end-to-end feedback from downstream tasks to learn an optimal masking strategy during pretraining. Our experimental findings highlight MLO-MAE's significant advancements in visual representation learning. Compared to existing methods, it demonstrates remarkable improvements across diverse datasets and tasks, showcasing its adaptability and efficiency. Our code is available at https://github.com/Alexiland/MLO-MAE 

ABSTRACT A comprehensive pangenomic approach was employed to analyze the genomes of 75 type II methylotrophs spanning various genera. Our investigation revealed 256 exact core gene families shared by all 75 organisms, emphasizing their crucial role in the survival and adaptability of these organisms. Additionally, we predicted the functionality of 12 hypothetical proteins. The analysis unveiled a diverse array of genes associated with key metabolic pathways, including methane, serine, glyoxylate, and ethylmalonyl-CoA (EMC) metabolic pathways. While all selected organisms possessed essential genes for the serine pathway,Methylooceanibacter marginalislacked serine hydroxymethyltransferase (SHMT), andMethylobacterium variabileexhibited both isozymes of SHMT, suggesting its potential to utilize a broader range of carbon sources. Notably,Methylobrevissp. displayed a unique serine-glyoxylate transaminase isozyme not found in other organisms. Only nine organisms featured anaplerotic enzymes (isocitrate lyase and malate synthase) for the glyoxylate pathway, with the rest following the EMC pathway.Methylovirgulasp. 4MZ18 stood out by acquiring genes from both glyoxylate and EMC pathways, andMethylocapsasp. S129 featured an A-form malate synthase, unlike the G-form found in the remaining organisms. Our findings also revealed distinct phylogenetic relationships and clustering patterns among type II methylotrophs, leading to the proposal of a separate genus forMethylovirgulasp. 4M-Z18 andMethylocapsasp. S129. This pangenomic study unveils remarkable metabolic diversity, unique gene characteristics, and distinct clustering patterns of type II methylotrophs, providing valuable insights for future carbon sequestration and biotechnological applications. IMPORTANCEMethylotrophs have played a significant role in methane-based product production for many years. However, a comprehensive investigation into the diverse genetic architectures across different genera of methylotrophs has been lacking. This study fills this knowledge gap by enhancing our understanding of core hypothetical proteins and unique enzymes involved in methane oxidation, serine, glyoxylate, and ethylmalonyl-CoA pathways. These findings provide a valuable reference for researchers working with other methylotrophic species. Furthermore, this study not only unveils distinctive gene characteristics and phylogenetic relationships but also suggests a reclassification forMethylovirgulasp. 4M-Z18 andMethylocapsasp. S129 into separate genera due to their unique attributes within their respective genus. Leveraging the synergies among various methylotrophic organisms, the scientific community can potentially optimize metabolite production, increasing the yield of desired end products and overall productivity. 

Noncoding RNAs (ncRNAs) play key roles in the regulation of important pathways, including cellular growth, stress management, signaling, and biofilm formation. Sulfate-reducing bacteria (SRB) contribute to huge economic losses causing microbial-induced corrosion through biofilms on metal surfaces. To effectively combat the challenges posed by SRB, it is essential to understand their molecular mechanisms of biofilm formation. This study aimed to identify ncRNAs in the genome of a model SRB, Oleidesulfovibrio alaskensis G20 (OA G20). Three in silico approaches revealed genome-wide distribution of 37 ncRNAs excluding tRNAs in the OA G20. These ncRNAs belonged to 18 different Rfam families. This study identified riboswitches, sRNAs, RNP, and SRP. The analysis revealed that these ncRNAs could play key roles in the regulation of several pathways of biosynthesis and transport involved in biofilm formation by OA G20. Three sRNAs, Pseudomonas P10, Hammerhead type II, and sX4, which were found in OA G20, are rare and their roles have not been determined in SRB. These results suggest that applying various computational methods could enrich the results and lead to the discovery of additional novel ncRNAs, which could lead to understanding the “rules of life of OA G20” during biofilm formation. 

Data literacy is increasingly relevant to everyday life and is a priority for educators across disciplinary boundaries. This study introduces a framework for characterizing data literacy instrucFon along five key dimensions. It then applies this framework to examine instances of data literacy instrucFon like explanaFons of data-related concepts and tasks/quesFons that invite learners to acFvely engage in data-related pracFces in a sample of lessons from a science and social studies high school textbook. By juxtaposing findings from science and social studies contexts, it examines how these disciplinary approaches compare with each other and idenFfies areas where these approaches could expand and build on each other to support more effecFve and holisFc data literacy development. 

In the present study, a thermophilic strain designated CamBx3 was isolated from the Campanario hot spring, Chile. Based on 16S rRNA gene sequence, phylogenomic, and average nucleotide identity analysis the strain CamBx3 was identified asBacillus paralicheniformis. Genome analysis ofB. paralicheniformisCamBx3 revealed the presence of genes related to heat tolerance, exopolysaccharides (EPS), dissimilatory nitrate reduction, and assimilatory sulfate reduction. The pangenome analysis of strain CamBx3 with eightBacillusspp. resulted in 26,562 gene clusters, 7,002 shell genes, and 19,484 cloud genes. The EPS produced byB. paralicheniformisCamBx3 was extracted, partially purified, and evaluated for its functional activities.B. paralicheniformisCamBx3 EPS with concentration 5 mg mL⁻¹showed an optimum 92 mM ferrous equivalent FRAP activity, while the same concentration showed a maximum 91% of Fe²⁺chelating activity.B. paralicheniformisCamBx3 EPS (0.2 mg mL⁻¹) demonstratedβ-glucosidase inhibition. The EPS formed a viscoelastic gel at 45°C with a maximum instantaneous viscosity of 315 Pa.s at acidic pH 5. The present study suggests thatB. paralicheniformisCamBx3 could be a valuable resource for biopolymers and bioactive molecules for industrial applications. 

Over the past decade, copper (Cu) has been recognized as a crucial metal in the differential expression of soluble (sMMO) and particulate (pMMO) forms of methane monooxygenase (MMO) through a mechanism referred to as the “Cu switch”. In this study, we used Methylosinus trichosporium OB3b as a model bacterium to investigate the range of Cu concentrations that trigger the expression of sMMO to pMMO and its effect on growth and methane oxidation. The Cu switch was found to be regulated within Cu concentrations from 3 to 5 µM, with a strict increase in the methane consumption rates from 3.09 to 3.85 µM occurring on the 6th day. Our findings indicate that there was a decrease in the fold changes in the expression of methanobactin (Mbn) synthesis gene (mbnA) with a higher Cu concentration, whereas the Ton-B siderophore receptor gene (mbnT) showed upregulation at all Cu concentrations. Furthermore, the upregulation of the di-heme enzyme at concentrations above 5 µM Cu may play a crucial role in the copper switch by increasing oxygen consumption; however, the role has yet not been elucidated. We developed a quantitative assay based on the naphthalene–Molisch principle to distinguish between the sMMO- and pMMO-expressing cells, which coincided with the regulation profile of the sMMO and pMMO genes. At 0 and 3 µM Cu, the naphthol concentration was higher (8.1 and 4.2 µM, respectively) and gradually decreased to 0 µM naphthol when pMMO was expressed and acted as the sole methane oxidizer at concentrations above 5 µM Cu. Using physical protein–protein interaction, we identified seven transporters, three cell wall biosynthesis or degradation proteins, Cu resistance operon proteins, and 18 hypothetical proteins that may be involved in Cu toxicity and homeostasis. These findings shed light on the key regulatory genes of the Cu switch that will have potential implications for bioremediation and biotechnology applications. 

Natural polysaccharides being investigated for use in the field of drug delivery commonly require the addition of sugars or pretreated biomass for fabrication. Geobacillus sp. strain WSUCF1 is a thermophile capable of secreting natural polymers, termed exopolysaccharides (EPSs), cultivated from cost-effective, non-treated lignocellulosic biomass carbon substrates. This preliminary investigation explores the capabilities of a 5% wt/wt amikacin-loaded film constructed from the crude EPS extracted from the strain WSUCF1. Film samples were seen to be non-cytotoxic to human keratinocytes and human skin-tissue fibroblasts, maintaining cell viability, on average, above 85% for keratinocytes over 72-h during a cell viability assay. The drug release profile of a whole film sample revealed a steady release of the antibiotic up to 12 h. The amikacin eluted by the EPS film was seen to be active against Staphylococcus aureus, maintaining above a 91% growth inhibition over a period of 48 h. Overall, this study demonstrates that a 5% amikacin-EPS film, grown from lignocellulosic biomass, can be a viable option for preventing or combating infections in clinical treatment. 

Scanning electron microscopy (SEM) techniques have been extensively performed to image and study bacterial cells with high-resolution images. Bacterial image segmentation in SEM images is an essential task to distinguish an object of interest and its specific region. These segmentation results can then be used to retrieve quantitative measures (e.g., cell length, area, cell density) for the accurate decision-making process of obtaining cellular objects. However, the complexity of the bacterial segmentation task is a barrier, as the intensity and texture of foreground and background are similar, and also, most clustered bacterial cells in images are partially overlapping with each other. The traditional approaches for identifying cell regions in microscopy images are labor intensive and heavily dependent on the professional knowledge of researchers. To mitigate the aforementioned challenges, in this study, we tested a U-Net-based semantic segmentation architecture followed by a post-processing step of morphological over-segmentation resolution to achieve accurate cell segmentation of SEM-acquired images of bacterial cells grown in a rotary culture system. The approach showed an 89.52% Dice similarity score on bacterial cell segmentation with lower segmentation error rates, validated over several cell overlapping object segmentation approaches with significant performance improvement. 

Cyber-physical systems are starting to adopt neural network (NN) models for a variety of smart sensing applications. While several efforts seek better NN architectures for system performance improvement, few attempts have been made to study the deployment of these systems in the field. Proper deployment of these systems is critical to achieving ideal performance, but the current practice is largely empirical via trials and errors, lacking a measure of quality. Sensing quality should reflect the impact on the performance of NN models that drive machine perception tasks. However, traditional approaches either evaluate statistical difference that exists objectively, or model the quality subjectively via human perception. In this work, we propose an efficient sensing quality measure requiring limited data samples using smart voice sensing system as an example. We adopt recent techniques in uncertainty evaluation for NN to estimate audio sensing quality. Intuitively, a deployment at better sensing location should lead to less uncertainty in NN predictions. We design SQEE, Sensing Quality Evaluation at the Edge for NN models, which constructs a model ensemble through Monte-Carlo dropout and estimates posterior total uncertainty via average conditional entropy. We collected data from three indoor environments, with a total of 148 transmitting-receiving (t-r) locations experimented and more than 7,000 examples tested. SQEE achieves the best performance in terms of the top-1 ranking accuracy---whether the measure finds the best spot for deployment, in comparison with other uncertainty strategies. We implemented SQEE on a ReSpeaker to study SQEE's real-world efficacy. Experimental result shows that SQEE can effectively evaluate the data collected from each t-r location pair within 30 seconds and achieve an average top-3 ranking accuracy of over 94%. We further discuss generalization of our framework to other sensing schemes.