ABSTRACT
The jelly from achenes of Ficus pumila var. awkeotsang (FPAA) is a famous beverage ingredient in Taiwan. In this work, ficumarin (1), a new compound was obtained from its twigs (FPAT) and elucidated with comprehensive spectroscopic data. The biosynthetic origin was proposed from the p-coumaroyl-CoA pathway. Alloxanthoxyletin, betulinic acid, and catechin were identified as the major and active constituents responsible for relieving neutrophilic inflammation by FPAT. Among them, the most potent alloxanthoxyletin was found to interact with PRO350 and GLU377 of human INOSOX. Further, Nrf2 activating capacity of the FPAT fraction and its coumarins was confirmed. With the analysis of LC-MS/MS data and feature-based molecular networking, coumarins were found as the dominant and responsible components. Notably, alloxanthoxyletin increased Nrf2 expression by up to 816.8 +/- 58% due to the interacting with the VAL561, THR560 and VAL420 residues of 5FNQ protein. COVID-19 Docking Server simulation indicated that pyranocoumarins would promisingly interfere with the life cycle of SARS-CoV-2. FPAT was proven to exert. Copyright © 2022 Taiwan Food and Drug Administration.
ABSTRACT
The trend of X-ray films recognition will become more and more important. This paper mainly uses neural networks to predict the symptoms of lung X-ray films of real patients for training and builds a user interface to facilitate image prediction. In addition to training and predicting the network, the core of the neural network will be implemented using a chip to speed up the computation that includes Winograd convolution, Max-pooling and Flatten operation. According to the experimental results, our method has a recognition rate of 88.1% for Covid-19. It can be seen that the method we propose has a high recognition rate for identifying Covid-19 and other pneumonia diseases. In addition, the operation speed of the proposed Winograd architecture is twice as fast as the traditional method. © 2021 IEEE.