Search

A novel image encryption algorithm using the random fractional DCT (RFrDCT) and the chaos-based Game of Life (GoL) is presented here. Firstly the plaintext image is transformed by the reality-prese...

Additional file 1: Table S1. Primers used in this experiment. Fig. S1. Plasmids constructed in this study. A DH overexpression plasmid. B ER overexpression plasmid. C ORFC heterologous plasmid. Fig. S2. ORFC heterologous expression in S. cerevisia YSG50. A The ORFC heterologous expression strain was selected by the Ura-depletion plate. B Genomic PCR analysis of ORFC. WT indicated the wild-type strain; YSG50-C indicated the ORFC heterologous strain. Fig. S3. A Genomic PCR products of DH and ER domains in the wild-type strain. B Plasmids construction validation by NotI and ApaI digestion. p-ER: ER-overexpression plasmid; p-DH: DH-overexpression plasmid. C Genomic PCR of Zeo expression cassette. P indicates positive control, N indicates the wild-type strain, DH+ indicates DH-overexpression strain, ER+ indicates ER-overexpression strain. Fig. S4 Gene copies of A DH and B ER domain in the wild-type and engineered strain, respectively, by qPCR analysis. All data are expressed as mean ± SD of three independent experiments. Fig. S5. Permutation test for the OPLS-DA model: A wild-type strain and DH-overexpressed strain and B wild-type strain and ER-overexpressed strain. Fig. S6. Metabolomics profiling by GC-MS reveals divergent metabolic phenotypes. A DH-overexpression strain compared with the wild-type strain. B ER-overexpression strain compared with the wild-type strain. Each node denotes an identified metabolite (red, up-regulated; blue, down-regulated; p < 0.05 by a two-tailed Student’s t-test). Node size reflects median fold change. Fig. S7. Heatmap of significant genes (log2 (fold change)) and their enriched pathways compared to the wild-type strain. Left frame: DH-overexpression strain; right frame: ER-overexpression strain. Table S2. Fatty acids composition analysis of domains located in ORFC overexpression strains

Deep learning has been used to image compressive sensing (CS) for enhanced reconstruction performance. However, most existing deep learning methods train different models for different subsampling ratios, which brings additional hardware burden. In this paper, we develop a general framework named scalable deep compressive sensing (SDCS) for the scalable sampling and reconstruction (SSR) of all existing end-to-end-trained models. In the proposed way, images are measured and initialized linearly. Two sampling masks are introduced to flexibly control the subsampling ratios used in sampling and reconstruction, respectively. To make the reconstruction model adapt to any subsampling ratio, a training strategy dubbed scalable training is developed. In scalable training, the model is trained with the sampling matrix and the initialization matrix at various subsampling ratios by integrating different sampling matrix masks. Experimental results show that models with SDCS can achieve SSR without changing their structure while maintaining good performance, and SDCS outperforms other SSR methods.

Strongly swirl flow simulation are still under developing. In this paper, e equation based on the Renormalization Group theory is used into algebraic stress model. Standard κ-e model, algebraic stress model by Jiang Zhang[5] and present model (RNG-ASM) are applied simultaneously to simulating the confined strongly swirling flow. The Simulating results by RNG-ASM model are compared to the results by other two model, it is shown that the predictions by this model display reasonable agreement with experimental data, and lead to greater improvement than Zhang’s ASM turbulence model[5].