芝麻是八大类食物过敏原之一,快速准确识别芝麻过敏原对预防其过敏有重要意义。核酸适配体可以高效识别靶标过敏原,在过敏原检测中有良好的应用前景。为了获得芝麻主要过敏原Ses i 2的特异性核酸适体,本研究以Ses i 2为靶标,通过磁珠筛...芝麻是八大类食物过敏原之一,快速准确识别芝麻过敏原对预防其过敏有重要意义。核酸适配体可以高效识别靶标过敏原,在过敏原检测中有良好的应用前景。为了获得芝麻主要过敏原Ses i 2的特异性核酸适体,本研究以Ses i 2为靶标,通过磁珠筛选法(磁珠-SELEX)开展10轮筛选,经由高通量测序获得6条候补序列(S1~S6),并进行家族性、同源性分析及二级结构预测。结果表明,6条候选核酸适体的重复率可达46.38%,其自由能在-9.02到-2.47 kcal·moL^(-1)之间,根据自由能能量稳定原则,S1和S5吉布斯自由能最低最稳定,分别为-6.70和-9.02 kcal·moL^(-1)。利用ELISA试验进行亲和力测试,结果表明核酸适体S1和S2的亲和能力较强,S1:KD=67.02 nmol·L^(-1),R2=0.925 8,S2:KD=97.65 nmol·L^(-1),R2=0.795 1。核酸适体S1与过敏原Ses i 2的结合力和其他过敏原蛋白相比有显著差异,可视为具有特异性。本研究最终获得一条兼具良好亲和力和特异性的核酸适体S1,为芝麻过敏原快速检测提供了技术支撑。展开更多
Typical sediments from Taihu Lake, a meso-to-hypereutrophic lake, were collected and examined on the basis of P-fractionation by sequential extraction scheme. Sedimentary inorganic phosphorus were fractioned into four...Typical sediments from Taihu Lake, a meso-to-hypereutrophic lake, were collected and examined on the basis of P-fractionation by sequential extraction scheme. Sedimentary inorganic phosphorus were fractioned into four forms and the rank order according to the mean concentration of P-fractions in Taihu Lake was NaOH-P>BD-P>HCl-P>NH_4Cl-P. The concentrations of BD-P were linearly correlated with the content of active Fe(R2=0.96). Also, the linear relationship between the sum of BD-P and NaOH-P and the sum of active Fe and active Al content was observed within the six sediments investigated(R2=0.96). Moreover, the bio-available phosphorus(BAP) content was estimated by the sum of NH_4Cl-P, BD-P, and NaOH-P, viz. BAP=NH_4Cl-P+NaOH-P+BD-P. In Taihu Lake, the BAP contents are ranging from 0.10 mg/g dw to 1.25 mg/g dw, and average 0.40 mg/g dw for all sediment samples. The relative contributions of BAP to total sedimentary phosphorus(TP) and inorganic sedimentary phosphorus(IP) range from 18.67% to 50.79%(33.61% on average) and from 52.82% to 82.09%(67.81% on average), respectively.展开更多
为了解决牛只多尺度、多目标、多部位状态信息的快速准确识别问题,笔者提出一种基于实时多目标注意力模型的牛只多部位快速自动化检测方法,采用多尺度注意力网络MSEAY,在网络中加入了SE(squeeze and excitation networks)注意力机制、H-...为了解决牛只多尺度、多目标、多部位状态信息的快速准确识别问题,笔者提出一种基于实时多目标注意力模型的牛只多部位快速自动化检测方法,采用多尺度注意力网络MSEAY,在网络中加入了SE(squeeze and excitation networks)注意力机制、H-swish激活函数及SIoU损失函数,消除现有牛群图像检测中光照、重叠、多类型等对图像检测结果产生的影响,先对牛只图像进行标注及预处理,然后采用MoibleNetv3模型作为Backbone进行牛只的部位划分识别,进而实现了多尺度、多部位牛只状态信息的精准分类与识别。结果表明:在选用含肉牛样本集合的前提下,各部位平均分类结果的精确率为96.6%,召回率为94.7%,F1值为94.1%,mAP@50值为97.4%;在仅选用种牛样本集合的前提下,各部位检测精确率为头部96.4%,躯干97.2%,腿部97.3%,平均值97.0%。本算法在保证检测实时性的同时提高了多尺度、多牛只的检测精度,验证了方法的鲁棒性与可泛化性,可用于多尺度、多目标、多部位状态信息的快速准确识别。展开更多
文摘芝麻是八大类食物过敏原之一,快速准确识别芝麻过敏原对预防其过敏有重要意义。核酸适配体可以高效识别靶标过敏原,在过敏原检测中有良好的应用前景。为了获得芝麻主要过敏原Ses i 2的特异性核酸适体,本研究以Ses i 2为靶标,通过磁珠筛选法(磁珠-SELEX)开展10轮筛选,经由高通量测序获得6条候补序列(S1~S6),并进行家族性、同源性分析及二级结构预测。结果表明,6条候选核酸适体的重复率可达46.38%,其自由能在-9.02到-2.47 kcal·moL^(-1)之间,根据自由能能量稳定原则,S1和S5吉布斯自由能最低最稳定,分别为-6.70和-9.02 kcal·moL^(-1)。利用ELISA试验进行亲和力测试,结果表明核酸适体S1和S2的亲和能力较强,S1:KD=67.02 nmol·L^(-1),R2=0.925 8,S2:KD=97.65 nmol·L^(-1),R2=0.795 1。核酸适体S1与过敏原Ses i 2的结合力和其他过敏原蛋白相比有显著差异,可视为具有特异性。本研究最终获得一条兼具良好亲和力和特异性的核酸适体S1,为芝麻过敏原快速检测提供了技术支撑。
基金The Grand Projects of Innovation Engineering Chinese Academy of Sciences(No. KZCX1 SW 12 2002 2006)
文摘Typical sediments from Taihu Lake, a meso-to-hypereutrophic lake, were collected and examined on the basis of P-fractionation by sequential extraction scheme. Sedimentary inorganic phosphorus were fractioned into four forms and the rank order according to the mean concentration of P-fractions in Taihu Lake was NaOH-P>BD-P>HCl-P>NH_4Cl-P. The concentrations of BD-P were linearly correlated with the content of active Fe(R2=0.96). Also, the linear relationship between the sum of BD-P and NaOH-P and the sum of active Fe and active Al content was observed within the six sediments investigated(R2=0.96). Moreover, the bio-available phosphorus(BAP) content was estimated by the sum of NH_4Cl-P, BD-P, and NaOH-P, viz. BAP=NH_4Cl-P+NaOH-P+BD-P. In Taihu Lake, the BAP contents are ranging from 0.10 mg/g dw to 1.25 mg/g dw, and average 0.40 mg/g dw for all sediment samples. The relative contributions of BAP to total sedimentary phosphorus(TP) and inorganic sedimentary phosphorus(IP) range from 18.67% to 50.79%(33.61% on average) and from 52.82% to 82.09%(67.81% on average), respectively.
文摘为了解决牛只多尺度、多目标、多部位状态信息的快速准确识别问题,笔者提出一种基于实时多目标注意力模型的牛只多部位快速自动化检测方法,采用多尺度注意力网络MSEAY,在网络中加入了SE(squeeze and excitation networks)注意力机制、H-swish激活函数及SIoU损失函数,消除现有牛群图像检测中光照、重叠、多类型等对图像检测结果产生的影响,先对牛只图像进行标注及预处理,然后采用MoibleNetv3模型作为Backbone进行牛只的部位划分识别,进而实现了多尺度、多部位牛只状态信息的精准分类与识别。结果表明:在选用含肉牛样本集合的前提下,各部位平均分类结果的精确率为96.6%,召回率为94.7%,F1值为94.1%,mAP@50值为97.4%;在仅选用种牛样本集合的前提下,各部位检测精确率为头部96.4%,躯干97.2%,腿部97.3%,平均值97.0%。本算法在保证检测实时性的同时提高了多尺度、多牛只的检测精度,验证了方法的鲁棒性与可泛化性,可用于多尺度、多目标、多部位状态信息的快速准确识别。
