连续时序的叶面积指数(Leaf Area Index,LAI)可反映苜蓿长势的变化情况,预测苜蓿未来时段的LAI对指导田间管理决策具有重要作用。针对LAI数据采集困难,导致苜蓿时序LAI存在训练数据不足的问题,该文以生长天数为自变量,采用修正的Logisti...连续时序的叶面积指数(Leaf Area Index,LAI)可反映苜蓿长势的变化情况,预测苜蓿未来时段的LAI对指导田间管理决策具有重要作用。针对LAI数据采集困难,导致苜蓿时序LAI存在训练数据不足的问题,该文以生长天数为自变量,采用修正的Logistic模型对实测苜蓿LAI变化的动态过程进行建模,根据LAI模拟曲线进行数据插补,从而构建宁夏引黄灌区试验区3年的逐日苜蓿LAI数据集。在插补数据集的基础上,为解决苜蓿刈割后数据突变问题,提出了一种ME-BiLSTM模型。该模型集成移动累计和检验方法(MOSUM)以及基于双向长短期记忆网络(BiLSTM)的编码器-解码器神经网络。MOSUM方法可以实现LAI数据集中突变点检测,并剔除包含突变点训练批次,同时应用改进的BiLSTM模型进行预测。结果表明:ME-BiLSTM模型能较好地进行苜蓿LAI未来曲线变化的预测,其决定系数(R^(2))、均方根误差(RMSE)值分别为0.9985和0.0722。对于苜蓿生长的各个茬次,预测模型对于第1茬、第4茬的预测精度最高,第2茬和第3茬的预测精度稍有降低。展开更多
Targeting the N gene of rabies virus (RV), four shRNA expression plasmids were designed and constructed based on the vector pRNATU6.3-Hygro that expresses fusion protein with GFP as a reporter gene. Four cell strains ...Targeting the N gene of rabies virus (RV), four shRNA expression plasmids were designed and constructed based on the vector pRNATU6.3-Hygro that expresses fusion protein with GFP as a reporter gene. Four cell strains (N1, N2, N3, N4) expressing the short hairpin RNAs (shRNA) were obtained after the plasmids were transfected into BHK-21 cells and screened under the pressure of Hygromycin B (300 μg/mL). These cell strains were infected with 100× the TCID 50 of rabies virus CVS-11 strain, and the viral replication was quantified at 24, 48, 72 and 96 hours by directed immunofluorescence assay (DFA), real-time PCR, and the 50% tissue culture infective dose (TCID 50 ). The results showed variable inhibition of viral replication, with BHK-N2 being the most effective strain (99% inhibition). There was close correspondence between results using the three methods of evaluation. The shRNA-mediated inhibition persisted to at least 96 hours after infection. Effective inhibition of replication of RV in BHK-21 cells was achieved by siRNA targeting the N gene, with N 2 , aimed at the region starting at position 701 of the gene, being the most potent.展开更多
文摘连续时序的叶面积指数(Leaf Area Index,LAI)可反映苜蓿长势的变化情况,预测苜蓿未来时段的LAI对指导田间管理决策具有重要作用。针对LAI数据采集困难,导致苜蓿时序LAI存在训练数据不足的问题,该文以生长天数为自变量,采用修正的Logistic模型对实测苜蓿LAI变化的动态过程进行建模,根据LAI模拟曲线进行数据插补,从而构建宁夏引黄灌区试验区3年的逐日苜蓿LAI数据集。在插补数据集的基础上,为解决苜蓿刈割后数据突变问题,提出了一种ME-BiLSTM模型。该模型集成移动累计和检验方法(MOSUM)以及基于双向长短期记忆网络(BiLSTM)的编码器-解码器神经网络。MOSUM方法可以实现LAI数据集中突变点检测,并剔除包含突变点训练批次,同时应用改进的BiLSTM模型进行预测。结果表明:ME-BiLSTM模型能较好地进行苜蓿LAI未来曲线变化的预测,其决定系数(R^(2))、均方根误差(RMSE)值分别为0.9985和0.0722。对于苜蓿生长的各个茬次,预测模型对于第1茬、第4茬的预测精度最高,第2茬和第3茬的预测精度稍有降低。
基金Project(2013CB632603)supported by the National Basic Research Program of ChinaProject(2015BAB19B02)supported by the National Key Technology R&D Program of ChinaProjects(51674084,51174051,51574082)supported by National Natural Science Foundation of China
文摘Targeting the N gene of rabies virus (RV), four shRNA expression plasmids were designed and constructed based on the vector pRNATU6.3-Hygro that expresses fusion protein with GFP as a reporter gene. Four cell strains (N1, N2, N3, N4) expressing the short hairpin RNAs (shRNA) were obtained after the plasmids were transfected into BHK-21 cells and screened under the pressure of Hygromycin B (300 μg/mL). These cell strains were infected with 100× the TCID 50 of rabies virus CVS-11 strain, and the viral replication was quantified at 24, 48, 72 and 96 hours by directed immunofluorescence assay (DFA), real-time PCR, and the 50% tissue culture infective dose (TCID 50 ). The results showed variable inhibition of viral replication, with BHK-N2 being the most effective strain (99% inhibition). There was close correspondence between results using the three methods of evaluation. The shRNA-mediated inhibition persisted to at least 96 hours after infection. Effective inhibition of replication of RV in BHK-21 cells was achieved by siRNA targeting the N gene, with N 2 , aimed at the region starting at position 701 of the gene, being the most potent.