Red-fleshed fruits are valued for their vibrant color and high anthocyanin content.Pre-harvest fruit bagging enhances fruit peel pigmentation,but its effect on flesh coloration remains poorly characterized.This study ...Red-fleshed fruits are valued for their vibrant color and high anthocyanin content.Pre-harvest fruit bagging enhances fruit peel pigmentation,but its effect on flesh coloration remains poorly characterized.This study revealed that removing bags from‘Gengcunyangtao’red-fleshed peach fruits triggers the rapid and uniform accumulation of anthocyanins in the flesh,resulting in anthocyanin levels that exceed those in unbagged fruits.The exposure to light after bag removal triggered significant increases in anthocyanin levels within 24 h.This was accompanied by the rapid upregulation of light-responsive and flavonoid biosynthetic gene expression levels within 6 h.A metabolomic analysis indicated that anthocyanin precursors,especially p-coumaric acid,accumulated before bag removal,thereby increasing substrate availability for rapid anthocyanin synthesis.On the basis of a weighted gene co-expression network analysis,MYB transcription factors,anthocyanin transporters,glutathione S-transferase,and multidrug and toxic compound extrusion(MATE)were identified as key regulators that coordinate precursor storage along with light-induced transcriptional activation.Notably,PpMYB4 binds to the promoter of PpGSTF14 and activates its expression,thereby promoting anthocyanin accumulation.The study findings elucidated the temporal coordination of metabolic priming and light-responsive transcriptional regulation driving rapid anthocyanin biosynthesis,with possible implications for improving peach fruit flesh coloration.展开更多
针对传统支持向量机(Support Vector Machine,SVM)集成学习(Ensemble Learning,EL)方法不能够解决高维复杂数据且子学习器差异性小集成效果不明显的问题,提出一种基于多种特征选择方法进行Bagging集成的支持向量机学习(Support Vector M...针对传统支持向量机(Support Vector Machine,SVM)集成学习(Ensemble Learning,EL)方法不能够解决高维复杂数据且子学习器差异性小集成效果不明显的问题,提出一种基于多种特征选择方法进行Bagging集成的支持向量机学习(Support Vector M achine Based on M ultiple Feature Selection Bagging,M FSB_SVM)方法.该方法首先采用不同的特征选择方法构建子学习器,以增加不同子学习器间的差异性,并直接从训练数据中对样本特征的重要性进行评估,而无需学习算法的反馈.实验表明,本文提出的MFSB_SVM方法既可以有效解决高维数据问题,也可避免传统SVM集成方法效果不明显的缺点,从而进一步提高学习模型的泛化性能.展开更多
基金supported by the Key Scientific and Technological Grant of Zhejiang for Breeding New Agricultural Varieties(Grant No.2021C12066-4)Huzhou Agricultural Science and Technology Innovation Team Project(Grant No.2022HN01).
文摘Red-fleshed fruits are valued for their vibrant color and high anthocyanin content.Pre-harvest fruit bagging enhances fruit peel pigmentation,but its effect on flesh coloration remains poorly characterized.This study revealed that removing bags from‘Gengcunyangtao’red-fleshed peach fruits triggers the rapid and uniform accumulation of anthocyanins in the flesh,resulting in anthocyanin levels that exceed those in unbagged fruits.The exposure to light after bag removal triggered significant increases in anthocyanin levels within 24 h.This was accompanied by the rapid upregulation of light-responsive and flavonoid biosynthetic gene expression levels within 6 h.A metabolomic analysis indicated that anthocyanin precursors,especially p-coumaric acid,accumulated before bag removal,thereby increasing substrate availability for rapid anthocyanin synthesis.On the basis of a weighted gene co-expression network analysis,MYB transcription factors,anthocyanin transporters,glutathione S-transferase,and multidrug and toxic compound extrusion(MATE)were identified as key regulators that coordinate precursor storage along with light-induced transcriptional activation.Notably,PpMYB4 binds to the promoter of PpGSTF14 and activates its expression,thereby promoting anthocyanin accumulation.The study findings elucidated the temporal coordination of metabolic priming and light-responsive transcriptional regulation driving rapid anthocyanin biosynthesis,with possible implications for improving peach fruit flesh coloration.
文摘针对传统支持向量机(Support Vector Machine,SVM)集成学习(Ensemble Learning,EL)方法不能够解决高维复杂数据且子学习器差异性小集成效果不明显的问题,提出一种基于多种特征选择方法进行Bagging集成的支持向量机学习(Support Vector M achine Based on M ultiple Feature Selection Bagging,M FSB_SVM)方法.该方法首先采用不同的特征选择方法构建子学习器,以增加不同子学习器间的差异性,并直接从训练数据中对样本特征的重要性进行评估,而无需学习算法的反馈.实验表明,本文提出的MFSB_SVM方法既可以有效解决高维数据问题,也可避免传统SVM集成方法效果不明显的缺点,从而进一步提高学习模型的泛化性能.
