[Effect of different data standardization methods on spectrum-effect relationship for anticoagulation of Trichosanthis Fructus dropping pills].

To study the effect of different data standardization methods on the spectrum-effect relationship for anticoagulant effect of Trichosanthis Fructus dropping pills. The spectrum-effect relationship was studied by using grey correlation degree method between three doses of Trichosanthis Fructus dropping pills and prothrombintime (PT) in mice. The effect of 10 data standardization methods, namely minimization method, maximum method, data extreme difference method, standard deviation standardization method, initialization transformation method, mean transformation method, ratio of each chromatographic peak area to the total peak area, ratio of each chromatographic peak area to the common peak area, logarithmic standardization method and tangent normalization method on the spectrum-effect relationship between Trichosanthis Fructus dropping pills and PT in mice was evaluated by using relative correlation degree as the index. The results of spectrum-effect relationship can be expressed by the minimization method, the data extreme difference method, the standard deviation standardization method, the initialization method and the mean transformation method, with highest relative correlation degree by the mean transformation method. As compared with the mean transformation method, there were significant differences between the high dose group and the medium dose group in the minimization method and the data extreme difference method (<0.01), while the minimization method in the low dose group showed statistical significance (<0.05). The standard deviation standardization method, initialization method and the mean transformation method can be used to study the spectrum-effect relationship for the anticoagulation of Trichosanthis Fructus dropping pills.