Learning to Sketch: A Neural Approach to Item Frequency Estimation in Streaming Data.

Recently, there has been a trend of designing neural data structures to go beyond handcrafted data structures by leveraging patterns of data distributions for better accuracy and adaptivity. Sketches are widely used data structures in real-time web analysis, network monitoring, and self-driving to estimate item frequencies of data streams within limited space. However, existing sketches have not fully exploited the patterns of the data stream distributions, making it challenging to tightly couple them with neural networks that excel at memorizing pattern information. Starting from the premise, we envision a pure neural data structure as a base sketch, which we term the meta-sketch, to reinvent the base structure of conventional sketches. The meta-sketch learns basic sketching abilities from meta-tasks constituted with synthetic datasets following Zipf distributions in the pre-training phase and can be quickly adapted to real (skewed) distributions in the adaption phase. The meta-sketch not only surpasses its competitors in sketching conventional data streams but also holds good potential in supporting more complex streaming data, such as multimedia and graph stream scenarios. Extensive experiments demonstrate the superiority of the meta-sketch and offer insights into its working mechanism.

Identification of potential biomarkers in Barrett's esophagus derived esophageal adenocarcinoma.

Almost 50% of esophageal adenocarcinoma (EAC) patients progressed from Barrett's esophagus (BE). EAC is often diagnosed at late stages and is related to dismal prognosis. However, there are still no effective methods for stratification and therapy in BE and EAC. Two public datasets (GSE26886 and GSE37200) were analyzed to identify differentially expressed genes (DEGs) between BE and EAC. Then, a series of bioinformatics analyses were performed to explore potential biomarkers associated with BE-EAC. 27 up- and 104 down-regulated genes were observed between GSE26886 and GSE37200. The GO and KEGG enrichment analysis indicated that the DEGs were highly involved in tumorigenesis. Subsequently, Weighted Gene Co-Expression Network Analysis (WGCNA) were performed to explore the potential genes related to BE-EAC, which were validated in The Cancer Genome Atlas (TCGA) database, and 5 up-regulated genes (MYO1A, ACE2, COL1A1, LGALS4, and ADRA2A) and 3 down-regulated genes (AADAC, RAB27A, and P2RY14) were found in EAC. Meanwhile, ADRA2A and AADAC could contribute to EAC pathogenesis and progression. MYO1A, ACE2, COL1A1, LGALS4, ADRA2A, AADAC, RAB27A, and P2RY14 could be potential novel diagnostic and prognostic biomarkers in BE-EAC.

A new pyroptosis-related signature for predicting the immune status and injury of liver ischemia-reperfusion.

OBJECTIVE: Pyroptosis is a type of programmed cell death. This study aimed to explore the roles of key pyroptosis-related genes in liver ischemia-reperfusion injury. METHODS: After collection and standardization of the transcriptome data from GSE12720 database, differentially expressed pyroptosis-related genes were identified. The risk genes screened by a random forest model were used to establish the line graph model. Consensus clustering was used to classify all samples according to the differentially expressed pyroptosis-related genes. Single-sample Gene Set Enrichment Analysis (ssGSEA) was performed to investigate the immune cell infiltration after hepatic ischemia-reperfusion. Cytoscape was used to visualize the regulatory network of transcription factor (TF)-microRNA (miRNA)-target genes. RESULTS: We identified 18 significantly and differentially expressed pyroptosis-related genes between the disease and normal samples. Among these 18 genes, IL1ß was positively correlated with CXCL8 (r = 0.791) and BIRC3 (r = 0.78), while ADORA3 was negatively correlated with GZMB (r = -0.567) and CXCL8 (r = -0.566). Furthermore, the random forest model constructed using the top 10 pyroptosis-related genes could predict the risk of hepatic ischemia-reperfusion. Importantly, the decision curve analysis showed that patients could benefit from the risk prediction model. Moreover, we found that the expression of TXNIP, IRF1, and GJA1 was the mostly regulated by miRNAs, while the expression of BIRC3, NFκB1, and TXNIP was regulated by the TF RELA. RELA had the most hub genes involved in the regulation. CONCLUSION: Our study provides an overview of the expression landscape and the functional significance of pyroptosis-related genes in liver ischemia-reperfusion. Our findings also shed light on the clinical application of pyroptosis-related genes in the treatment of hepatic ischemia-reperfusion injury.