Because traditional Chinese medicine (TCM) feature complicated active ingredients and diverse compatible herbs

it is difficult to define the active ingredients and their synergistic effect and even conduct a study on the compatibility between single TCM ingredients. The spectrum-effect relationship

as an emerging method

has been gradually applied in studies on the effect material foundation of compounds

component compatibility

processing mechanism

efficiency prediction and process optimization. The study method is a data-based analytical techniques to associate the multidimensional fingerprint data with the pharmacological indexes as well as a crossing field between pharmacy and mathematics. The study relies on the data analysis technique and focuses on selecting objective and reasonable data analysis technique. In this paper

the authors summarized the advance in applications of data analysis technique in the spectrum-effect relationship

including

forecasting the analysis method on the correlation between ingredients and efficacy:grey relational analysis

artificial neural networks

bivariate correlations analysis. Defining the analysis methods on the contribution of each component to the efficacy:multiple linear regression

partial least squares;simplifying the analysis method for finding major active ingredients:principal component analysis

canonical correlation analysis. Besides

the characteristics of currently common data analysis technique

research ideas and their applicable research objects were summarized. At last

the authors discussed the limitations in various data analysis technique

gave suggestions for selecting data analysis technique in the spectrum-effect relationship and recommended the reasonable combination of a variety of data analysis technique

so as to establish an ideal spectrum-effect relationship model and ensure the objectivity and authenticity of the results.