Search

The bicistronic design (BCD) is characterized by a short fore-cistron sequence and a second Shine-Dalgarno (SD2) sequence upstream of the target gene. The outstanding performance of this expression cassette in promoting recombinant protein production has attracted attention. Recently, the application of the BCD has been further extended to gene expression control, protein translation monitoring, and membrane protein production. In this review, we summarize the characteristics, molecular mechanisms, applications, and structural optimization of the BCD expression cassette. We also specifically discuss the challenges that the BCD system still faces. This is the first review of the BCD expression strategy, and it is believed that an in-depth understanding of the BCD will help researchers to better utilize and develop it. KEY POINTS: • Summary of the characteristics and molecular mechanisms of the BCD system. • Review of the actual applications of the BCD expression cassette. • Summary of the structural optimization of the BCD system.

Aiming at the lack of effective quantitative model to support the analysis of terrorist attacks, a multilayer depth Neural network (NN) Graph convolutional networks (GCN) model (NNGCN) was put forward to realize the classification and early warning of terrorist attacks. The proposed model optimized the traditional GCN with the help of complex NN. The concept of link index was introduced into the NNGCN model. It is combined with the important information between event nodes. The information includes the similarity of events and link probability. Compared with the original unoptimized model, the improved model increased the classification accuracy of terrorist attacks. Because the model uses the node's feature information and the link relationship of graph structure, it can also warn the sudden terrorist attacks effectively.

Outer membrane lipoprotein A (OmlA) is a vaccine antigen against porcine contagious pleuropneumonia (PCP), a disease severely affecting the swine industry. Here, we aimed to systematically potentiate the secretory production of OmlA in Corynebacterium glutamicum (C. glutamicum), a widely used microorganism in the food industry, by establishing a holistic development process based on our high-throughput culture platform. The expression patterns, expression element combinations, medium composition, and induction conditions were comprehensively screened or optimized in microwell plates (MWPs), followed by fermentation parameter optimization in a 4×1 L parallel fermentation system (CUBER4). An unprecedented yield of 1.01 g/L OmlA was ultimately achieved in a 5-L bioreactor following the scaling-up strategy of fixed oxygen mass transfer coefficient (kLa), and the produced OmlA antigen showed well-protective immunity against Actinobacillus pleuropneumoniae challenge. This result provides a rapid and reliable pipeline to achieve the hyper-production of OmlA, and possibly other recombinant vaccines, in C. glutamicum.

Session-based recommendation is used to generate a session recommendation list for the current user according to the existing session information. Most of the existing session-based recommendation models with deep learning only focus on the session information of users, rarely consider the interaction recommendation between similar users, and ignore the user's interest orientation, so the accuracy of the session-based recommendation still has space to improve. To solve above problems, an application model of GNN in session recommendation was proposed. Firstly, take the item's number and order accessed by user session as the basis of clustering, k-means algorithm was used to cluster the sessions to generate k similar sessions clusters. Secondly, choose the appropriate atomic interest on the basis of session clustering to expand user's session length and strengthen user's interest orientation. Thirdly, train the enhanced user sessions by using graph neural network to obtain the latent vector of items, and predict the probability of item being clicked according to the items and the latent vector of the given session, and then generate session recommendation list. Finally, the proposed model is compared with other session recommendation models in YOOCHOOSE and DIGINETICA dataset. The experimental results show that the proposed model has higher recommendation accuracy.

To study etiology of clinically diagnosed non A-E hepatitis.HBV, TTV, human parvovirus B19, SENV DNA were detected by nested polymerase chain reactions (nPCR), while HGV, HCV RNA were tested by reverse transcription nested polymerase chain reactions (RT-nPCR).Of 60 patients with clinically diagnosed non A-E hepatitis, 30 (50.0%) were HBV DNA positive alone, 10 (16.7%) HBV and TTV DNA positive, 6 (10.0%) HBV and B19 DNA positive; 1 (1.7%) HBV, SENV DNA and HCV RNA positive, 1 (1.7%) HCV RNA positive alone, 1 (1.7%) HCV RNA and B19 DNA positive, 2 (3.3%) B19 DNA positive alone, 1 (1.7%) TTV DNA positive alone, and the remaining 8 (13.3%) negative for all viruses. All the 60 patients were HGV RNA negative. There were no differences in serum biochemical markers of hepatitis B patients with or without TTV or B19 virus infection.HBV is a major etiologic agent for the clinically diagnosed non A-E hepatitis. HGV, TTV, B19 and SEBV may not be associated with nonA-E hepatitis.