Preparation, identification, activity prediction, and protective effects on IR-HepG2 cells of five novel DPP-IV inhibitory peptides from protein hydrolysate of skipjack tuna dark muscles.

To produce peptides with high dipeptidyl peptidase IV (DPP-IV) inhibitory activity, neutrase was selected from five proteases (trypsin, neutrase, pepsin, alcalase and flavor protease) with the highest degree of hydrolysis (DH) (18.23 ± 1.08%) and DPP-IV inhibitory rate (53.35 ± 4.02%) to produce protein hydrolysate (NPH) from the dark muscles of skipjack tuna (Katsuwonus pelamis). Then, NPH-1 was isolated from NPH by gel permeation chromatography and found to possess the highest DPP-IV inhibitory rate (65.12 ± 7.94% at 0.5 mg ml-1) in the separated components (including NPH-1, NPH-2, NPH-3 and NPH-4). Subsequently, the available prediction models of tripeptides and tetrapeptides with the DPP-IV inhibitory rate were established using an artificial neural network (ANN). The RMSE (0.56 and 0.33 for the model established through collected tripeptides and tetrapeptides, respectively) and R2 (0.95 and 0.99 for the model established through collected tripeptides and tetrapeptides, respectively) of the ANN model's parameters were within acceptable limits, indicating that this model is available. Next, the ANN model was applied to predict tripeptides and tetrapeptides from the hydrolysate of skipjack tuna dark muscles, and five peptides (Ala-Pro-Pro (APP), Pro-Pro-Pro (PPP), Asp-Pro-Leu-Leu (DPLL), Glu-Ala-Val-Pro (EAVP) and Glu-Ala-Iie-Pro (EAIP)) possessing a noticeable DPP-IV inhibitory rate (with DPP-IV IC50 values of 42.46 ± 5.02, 37.71 ± 9.17, 58.85 ± 14.42, 49.94 ± 6.69 and 57.15 ± 6.13 µM, respectively) were screened from the protein hydrolysate. The above five peptides were proved to effectively promote glucose consumption in the insulin resistant-HepG2 (IR-HepG2) cell model considering that the glucose consumption rates of APP, PPP, DPLL, EAVP and EAIP treatment groups are all more than twice that of the dexamethasone group. Accordingly, mechanistic studies showed that these peptides interacted with PI3K/AKT and AMPK signaling pathways and promoted the phosphorylation of PI3K p110, AKT and AMPK (the protein expressions of PI3K p110, p-AKT and p-AMPK in APP, PPP, DPLL, EAVP and EAIP treatment groups are 1.64-2.22 fold compared with that in the dexamethasone group), thereby enhancing glucose uptake and further alleviating insulin resistance. These findings demonstrated that skipjack tuna dark muscle is a potential DPP-IV inhibitory peptide source, and five DPP-IV inhibitory peptides from its hydrolysate may exert potent anti-diabetic activity. In comparison, PPP may be the most potential active ingredient for healthy food against type 2 diabetes mellitus in the five screened peptides considering synthetically the DPP-IV inhibitory rate, bioavailability and synthesis cost.

Using the meteorological early warning model to improve the prediction accuracy of water damage geological disasters around pipelines in mountainous areas.

This paper focuses on the threat of water damage geological disasters brought by the complex terrain along the long-distance natural gas pipeline. The role of rainfall factors in the occurrence of such disasters has been fully considered, a meteorological early warning model for water damage geological disasters in mountainous areas based on slope units has been constructed to improve the prediction accuracy of such disasters and timely early warning and forecasting. An actual natural gas pipeline in a typical mountainous area of Zhejiang Province is taken as an example. The hydrology-curvature combined analysis method is chosen to divide the slope units, and the SHALSTAB model is used to fit the slope soil environment to calculate the stability level. Finally, the stability level is coupled with rainfall data to calculate the early warning index for water damage geological disasters in the study area. The results show that compared with the separate SHALSTAB model, the early warning results coupled with rainfall are more effective in predicting water damage geological disasters. The early warning results are compared with the actual disaster points, among the nine actual disaster points, most of the slope units around seven disaster points are in the state of needing early warning, the early warning accuracy rate reaches 77.8 %. The proposed early warning model can carry out targeted deployment in advance according to the divided slope units, and the prediction accuracy of geological disasters induced by heavy rainfall weather is significantly higher and more suitable for the actual location of the disaster point, which can provide a basis for accurate disaster prevention in the research area and areas with similar geological environments.

Φ-OTDR Signal Identification Method Based on Multimodal Fusion.

Distributed fiber optic sensing (DFS) systems are an effective method for long-distance pipeline safety inspections. Highly accurate vibration signal identification is crucial to DFS. In this paper, we propose an end-to-end high-accuracy fiber optic vibration signal detection and identification algorithm by extracting features from the time domain and frequency domain by a one-dimensional convolutional neural network and two-dimensional convolutional neural network, respectively, and introducing a self-attentive mechanism to fuse the features of multiple modes. First, the raw signal is segmented and normalized according to the statistical characteristics of the vibration signal combined with the distribution of noise. Then, the one-dimensional sequence of vibration signal and its two-dimensional image generated by short-time Fourier transform are input to the one-dimensional convolutional neural network and two-dimensional neural network, respectively, for automatic feature extraction, and the features are combined by long and short-time memory. Finally, the multimodal features generated from the time and frequency domains are fused by a multilayer TransformerEncoder structure with a multiheaded self-attentive mechanism and fed into a multilayer perceptron for classification. Experiments were conducted on an urban field database with complex noise and achieved 98.54% accuracy, which demonstrates the effectiveness of the proposed algorithm.