Ultrasound-Assisted Preparation of Maillard Reaction Products Derived from Hydrolyzed Soybean Meal with Meaty Flavor in an Oil-In-Water System.

In the present work, we prepared Maillard reaction products (MRPs) derived from enzyme hydrolyzed soybean meal with ultrasound assistance in an oil-(oxidized lard)-in-water system (UEL-MRPs) or oil-free system (UN-MRPs), and the effect of ultrasound on the properties of the obtained MRPs was evaluated. The analysis of fatty acids in lard with different treatments showed that ultrasound can generate more unsaturated fatty acids in the aqueous phase. The UV-Vis absorbances of UEL-MRPs, UN-MRPs, and MRPs obtained in an oil-in-water system (EL-MRPs) and MRPs obtained in an oil-free system (N-MRPs) at 294 and 420 nm indicated that ultrasound could increase the amount of Maillard reaction intermediates and melanoids in the final products of the Maillard reaction. This was in line with the result obtained from color change determination-that ultrasound can darken the resultant MRPs. Volatile analysis showed ultrasound can not only increase the number of volatile substances, but also greatly increase the composition of volatile substances in UEL-MRPs and UN-MRPs, especially the composition of those contributing to the flavor of the MRPs, such as oxygen-containing heterocycles, sulfur-containing compounds, and nitrogen-containing heterocycles. Descriptive sensory evaluation revealed that UN-MRPs and UEL-MRPs had the highest scores in total acceptance, ranking in the top two, and UEL-MRPs had the strongest meaty flavor among these four kinds of MRPs. Furthermore, the measurements of antioxidant activities, including DPPH radical-scavenging activity, hydroxyl radical scavenging ability, and ferric ion reducing antioxidant power, were conducted, showing that UN-MRPs exhibited the highest antioxidant activity among all the MRPs.

Producing beef flavors in hydrolyzed soybean meal-based Maillard reaction products participated with beef tallow hydrolysates.

This work investigated the effect of oxidized beef tallow on the volatile compositions and sensory properties of soybean meal-based Maillard reaction products (MRPs). Various tallow oxidation methods included thermal treatment (TT), enzymatic hydrolysis (ET) and enzymatic hydrolysis combined with mild thermal (ETT) treatment. Results showed that all these oxidized tallow contained more types of volatile compounds than those of untreated tallow. Moreover, the composition of almost all types of volatile substances was greatly increased with the addition of the oxidized beef tallow into the hydrolyzed soybean meal-based Maillard reaction system. More importantly, the composition of oxygen-containing heterocycles (63.89 µg/mL), sulfur-containing compounds (76.64 µg/mL), and nitrogen-containing heterocycles (19.81 µg/mL) that contribute positively to sensory properties in ETT-MRPs was found to be the highest among all the MRPs. Correlation assessment revealed that ETT was closely related to the most typical volatile products and sensory attributes, indicating this approach can effectively enhance the sensory and flavor of hydrolyzed soybean meal derived MRPs.

Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma.

BACKGROUND: An accurate pathological diagnosis of hepatocellular carcinoma (HCC), one of the malignant tumors with the highest mortality rate, is time-consuming and heavily reliant on the experience of a pathologist. In this report, we proposed a deep learning model that required minimal noise reduction or manual annotation by an experienced pathologist for HCC diagnosis and classification. METHODS: We collected a whole-slide image of hematoxylin and eosin-stained pathological slides from 592 HCC patients at the First Affiliated Hospital, College of Medicine, Zhejiang University between 2015 and 2020. We propose a noise-specific deep learning model. The model was trained initially with 137 cases cropped into multiple-scaled datasets. Patch screening and dynamic label smoothing strategies are adopted to handle the histopathological liver image with noise annotation from the perspective of input and output. The model was then tested in an independent cohort of 455 cases with comparable tumor types and differentiations. RESULTS: Exhaustive experiments demonstrated that our two-step method achieved 87.81% pixel-level accuracy and 98.77% slide-level accuracy in the test dataset. Furthermore, the generalization performance of our model was also verified using The Cancer Genome Atlas dataset, which contains 157 HCC pathological slides, and achieved an accuracy of 87.90%. CONCLUSIONS: The noise-specific histopathological classification model of HCC based on deep learning is effective for the dataset with noisy annotation, and it significantly improved the pixel-level accuracy of the regular convolutional neural network (CNN) model. Moreover, the model also has an advantage in detecting well-differentiated HCC and microvascular invasion.

Parameter Estimation for Interrupted Sampling Repeater Jamming Based on ADMM.

By repeatedly sampling, storing, and retransmitting parts of the radar signal, interrupted sampling repeater jamming (ISRJ) based on digital radio frequency memory (DRFM) can produce a train of secondary false targets symmetrical to the main false target, threatening to mislead or deceive the victim radar system. This paper proposes a computationally-effective method to estimating the parameters for ISRJ by resorting to the framework of alternating direction method of multipliers (ADMM). Firstly, the analytical form of pulse compression is derived. Then, for the purpose of estimating the parameters of ISRJ, the original problem is transformed into a nonlinear integer optimization model with respect to a window vector. On this basis, the ADMM is introduced to decompose the nonlinear integer optimization model into a series of sub-problems to estimate the width and number of ISRJ's sample slices. Finally, the numerical simulation results show that, compared with the traditional time-frequency (TF) method, the proposed method exhibits much better performance in accuracy and stability.

Genome-wide DNA methylation changes associated with olfactory learning and memory in Apis mellifera.

The honeybee is a model organism for studying learning and memory formation and its underlying molecular mechanisms. While DNA methylation is well studied in caste differentiation, its role in learning and memory is not clear in honeybees. Here, we analyzed genome-wide DNA methylation changes during olfactory learning and memory process in A. mellifera using whole genome bisulfite sequencing (WGBS) method. A total of 853 significantly differentially methylated regions (DMRs) and 963 differentially methylated genes (DMGs) were identified. We discovered that 440 DMRs of 648 genes were hypermethylated and 274 DMRs of 336 genes were hypomethylated in trained group compared to untrained group. Of these DMGs, many are critical genes involved in learning and memory, such as Creb, GABA B R and Ip3k, indicating extensive involvement of DNA methylation in honeybee olfactory learning and memory process. Furthermore, key enzymes for histone methylation, RNA editing and miRNA processing also showed methylation changes during this process, implying that DNA methylation can affect learning and memory of honeybees by regulating other epigenetic modification processes.