Abstract:

The quantitative structure-retention index relationships (QSRR) of 56 volatile compounds in Houttuynia cordata Thunb were studied, using molecular connectivity indices, molecular shape indices and the electrotopological state indices to describe the chemical structure of the volatile compounds. Various multiple linear regression models were established with variable screening by the stepwise multiple regression technique and statistics. QSRR for estimating retention index (RI)of volatile compounds in Houttuynia cordata Thunb were achieved based on multiple regression, and the regression equations were proposed with the correlation coefficient 0.991. In order to test the stability and prediction ability of the model, the leave-oneout procedure was performed and good results (R ＞ 0.99)were obtained. These results indicate that molecular connectivity indices, molecular shape indices and the electrotopological state indices have good rationality and efficiency for attribute RI of volatile compounds of Houttuynia cordata Thunb.

Key words: Houttuynia cordata Thunb, retention index, volatile compounds, QSRR