Search

Deep learning adjusts parameters to optimize the model through model training. Training algorithm is the key to model optimization and implementation. Therefore, the improvement of model training algorithm is of great significance to deep learning. Based on the original gradient algorithm, the paper proposes a new gradient descent algorithm AdaGrad Restricted by Windows (AdaRW) for Deep learning model training optimization. Aiming at the defects of AdaGrad algorithm, the new algorithm uses the subset of window to limit historical accumulation, so as to slow down the attenuation of learning rate, and improve the speed of model training. The paper constructs OceanTDA9, a Deep learning model of marine target detection for Synthetic Aperture Radar (SAR) data, and adopts the proposed AdaRW algorithm to train the model based on SAR data with a resolution of 10 m in the Bohai Sea. Experiment shows that the accuracy and loss of the algorithm are better than those of AdaGrad and Stochastic Gradient Descent (SGD) algorithms, and the standard deviation is better than that of Adam algorithm. [ABSTRACT FROM AUTHOR]

Shigella is an intracellular pathogen that primarily infects the human colon and causes shigellosis. Shigella virulence relies largely on the type III secretion system (T3SS) and secreted effectors. VirF, the master Shigella virulence regulator, is essential for the expression of T3SS-related genes. In this study, we found that YhjC, a LysR-type transcriptional regulator, is required for Shigella virulence through activating the transcription of virF. Pathogenicity of the yhjC mutant, including colonization in the colons of guinea pigs as well as its ability for host cell adhesion and invasion, was significantly lowered. Expression levels of virF and nearly all VirF-dependent genes were downregulated by yhjC deletion, indicating that YhjC can activate virF transcription. Electrophoretic mobility shift assay analysis demonstrated that YhjC could bind directly to the virF promoter region. Therefore, YhjC is a novel virulence regulator that positively regulates the virF expression and promotes Shigella virulence. Additionally, genome-wide expression analysis identified the presence of other genes in the large virulence plasmid and a genome exhibiting differential expression in response to yhjC deletion, with 169 downregulated and 99 upregulated genes, indicating that YhjC also functioned as a global regulatory factor. [ABSTRACT FROM AUTHOR]

The intracellular pathogen Salmonella enterica serovar Typhimurium (S. Typhimurium) exploits host macrophage as a crucial survival and replicative niche. To minimize host immune response stimulated by flagellin, the expression of flagellar genes is downregulated during S. Typhimurium growth within host macrophages. However, the underlying mechanisms are largely unknown. In this study, we show that STM14_1285 (named AsiR), a putative RpiR-family transcriptional regulator, which is downregulated within macrophages as previously reported and also confirmed here, positively regulates the expression of flagellar genes by directly binding to the promoter of flhDC. By generating an asiR mutant strain and a strain that persistently expresses asiR gene within macrophages, we confirmed that the downregulation of asiR contributes positively to S. Typhimurium replication in macrophages and systemic infection in mice, which could be attributed to decreased flagellar gene expression and therefore reduced flagellin-stimulated secretion of pro-inflammatory cytokines IL-1β and TNF-α. Furthermore, the acidic pH in macrophages is identified as a signal for the downregulation of asiR and therefore flagellar genes. Collectively, our results reveal a novel acidic pH signal-mediated regulatory pathway that is utilized by S. Typhimurium to promote intracellular replication and systemic pathogenesis by repressing flagellar gene expression. [ABSTRACT FROM AUTHOR]