Deep clustering using 3D attention convolutional autoencoder for hyperspectral image analysis.

Deep clustering has been widely applicated in various fields, including natural image and language processing. However, when it is applied to hyperspectral image (HSI) processing, it encounters challenges due to high dimensionality of HSI and complex spatial-spectral characteristics. This study introduces a kind of deep clustering model specifically tailed for HSI analysis. To address the high dimensionality issue, redundant dimension of HSI is firstly eliminated by combining principal component analysis (PCA) with t-distributed stochastic neighbor embedding (t-SNE). The reduced dataset is then input into a three-dimensional attention convolutional autoencoder (3D-ACAE) to extract essential spatial-spectral features. The 3D-ACAE uses spatial-spectral attention mechanism to enhance captured features. Finally, these enhanced features pass through an embedding layer to create a compact data-representation, and the compact data-representation is divided into distinct clusters by clustering layer. Experimental results on three publicly available datasets validate the superiority of the proposed model for HSI analysis.

Length and rigidity of the spacer impact on aldose reductase inhibition of the 5F-like ARIs in a dual-occupied mode.

The primary objective of this study was to investigate the structure-activity relationship of a new series of 5F-like Aldose Reductase Inhibitors (ARIs) using in silico docking method. In this perspective, 6 novel ARIs have been designed and synthesized. Evaluation of the inhibition of these compounds to ALR2 was carried on with epalrestat and 5F as the references. It was found that the spacer of 5F-like ARIs has a great influence on their inhibitory activity. Rigid spacer with length equal to 3 âˆ¼ 4 carbon alkyl chain brings about better inhibitory activity. Among them, compound 4b was verified as the most active ARIs, where its IC50 value was 16.8 ± 1.3 nM. Furthermore, in silico docking studies using AutoDock 4.2 as well as molecular simulation using GROMACS 2022.1 showed that 5F-like ARIs adopt a dual-occupation mode. The interaction energy (-25 to -74 kcal/mol), as well as MM-GBSA binding free energy (-37 to -65 kcal/mol) was positively correlated with their ALR2 inhibition constant (2000 to 16.8 nM). Docking interaction explained well the structure-activity relationship. A pharmacophore model has been set up for 5F-like ARIs thereafter. This model indicates that as an effective ARI, the entity should have four characteristics: an aromatic center, two hydrogen bond donors, and one hydrogen bond acceptor. By the way, all the 5F-like ARIs reported here are good to mild antioxidant with EC50 value between 13.6 ± 1.2 and 71.1 ± 3.2 µM. All our data direct the further development of more optimal ARIs for the treatment of diabetic complication in the future.

SRSF1 promotes the inclusion of exon 3 of SRA1 and the invasion of hepatocellular carcinoma cells by interacting with exon 3 of SRA1pre-mRNA.

Steroid receptor RNA activator 1 (SRA1) has been described as a novel transcriptional co-activator that affects the migration of cancer cells. Through RT-PCR, we identified that skipping exon 3 of SRA1 produces two isoforms, including the truncated short isoform, SRA1-S, and the long isoform, SRA1-L. However, the effect of these two isomers on the migration of HCC cells, as well as the specific mechanism of exon 3 skipping remain unclear. In this study, we found up regulated expression of SRSF1 and SRA1-L in highly metastatic HCCLM3, as well as in HCCs with SRSF1 demonstrating the strongest correlation with SRA1-L. In contrast, we observed a constitutively low expression of SRA1-S and SRSF1 in lowly metastatic HepG2 cells. Overexpression of SRSF1 or SRA1-L promoted migration and invasion by increasing the expression of CD44, while SRA1-S reversed the effect of SRSF1 and SRA1-L in vitro. In addition, lung metastasis in mice revealed that, knockdown of SRSF1 or SRA1-L inhibited the migration of HCC cells, while SRA1-L overexpression abolished the effect of SRSF1 knockout and instead promoted HCC cells migration in vivo. More importantly, RNA immunoprecipitation and Cross-link immunoprecipitation analyses showed that SRSF1 interacts with exon 3 of SRA1 to up regulate the expression of SRA1-L in HCC cells. RNA pull-down results indicated that SRSF1 could also bind to exon 3 of SRA1 in vitro. Finally, minigene -MS2 mutation experiments showed that mutation of the SRA1 exon 3 binding site for SRSF1 prevented the binding of SRA1 pre-mRNA. In summary, our results provide experimental evidence that SRA1 exon 3 inclusion is up regulated by SRSF1 to promote tumor invasion and metastasis in hepatocellular carcinoma.