ε-Poly-l-lysine (ε-PL) is a natural antimicrobial polymer with significant inhibitory activity against a broad spectrum of microorganisms, and nowadays used widely as a preservative in the food industry. In the present study, ε-PL broth was obtained from Streptomyces ahygroscopicus GIM8 fermentation in a nutrient-limited liquid medium. The in vitro antifungal activity of the broth against fruit pathogens Penicillium expansum and Colletotrichum gloeosporioides was investigated, and its usage for postharvest storage of two highly perishable fruits wax apple and guava was evaluated. Results showed that ε-PL concentration in the broth reached 0.61 g/L, and the nutrition level of the broth was low. The antifungal activity of ε-PL broth was comparable to that of the aqueous solution of ε-PL under the same concentration. Immersion with the diluted broth (200 mg/L ε-PL) markedly delayed the decline in the quality of postharvest wax apple and guava fruits during storage, and the decay incidences were also greatly decreased as compared to their respective controls (distilled water immersion). A further investigation demonstrated that the ε-PL broth immersion induced an increase in the activity of defense-related enzymes peroxidase and polyphenol oxidase in the two fruits during storage. The present study proved that the fermentation broth of ε-PL could be used as a promising alternative to high purity ε-PL and synthetic fungicides for preserving fruits at postharvest stage.

IntroductionThe successful use of machine learning (ML) for medical diagnostic purposes has prompted myriad applications in cancer image analysis. Particularly for hepatocellular carcinoma (HCC) grading, there has been a surge of interest in ML-based selection of the discriminative features from high-dimensional magnetic resonance imaging (MRI) radiomics data. As one of the most commonly used ML-based selection methods, the least absolute shrinkage and selection operator (LASSO) has high discriminative power of the essential feature based on linear representation between input features and output labels. However, most LASSO methods directly explore the original training data rather than effectively exploiting the most informative features of radiomics data for HCC grading. To overcome this limitation, this study marks the first attempt to propose a feature selection method based on LASSO with dictionary learning, where a dictionary is learned from the training features, using the Fisher ratio to maximize the discriminative information in the feature.MethodsThis study proposes a LASSO method with dictionary learning to ensure the accuracy and discrimination of feature selection. Specifically, based on the Fisher ratio score, each radiomic feature is classified into two groups: the high-information and the low-information group. Then, a dictionary is learned through an optimal mapping matrix to enhance the high-information part and suppress the low discriminative information for the task of HCC grading. Finally, we select the most discrimination features according to the LASSO coefficients based on the learned dictionary.Results and discussionThe experimental results based on two classifiers (KNN and SVM) showed that the proposed method yielded accuracy gains, compared favorably with another 5 state-of-the-practice feature selection methods.