关键词: 同源基因, 同源基因

Abstract:

For studying the problem that forecasting flue-cured tobacco quality with meteorological factors, the chemical quality of flue-cured tobacco in 83 tobacco-planting counties of Yunnan Province in 2004 were divided into 10 types, which was based on total sugar, reducing sugar, nicotine, total nitrogen and protein, according to the principle of minimum square Euclidean distance by using systemic cluster analysis. According to the theory that similar chemical compositions of flue-cured tobacco in similar climatic conditions, the Euclidean distance of meteorological conditions between the 83 tobacco-planting counties in 2005 and the corresponding 10 types of chemical quality in flue-cured tobacco were calculated. To the principle that the smaller distance, the higher similarity, the main chemical compositions and quality types of tobacco leaf were predicted. The results showed that percent relative root mean square error (RRMSE) of total sugar, reducing sugar, nicotine, total nitrogen and protein predictive values were 8.91%, 10.40%, 22.31%, 10.46% and 13.08%, respectively. The predictive effect of this method was general better in mid-value and high-value areas than in low-value area of chemical compositions. The results suggested that this method has a certain feasibility for chemical compositions forecast of tobacco.