A fast method for predicting adenosine content in porcini mushrooms using Fourier transform near-infrared spectroscopy combined with regression model

Adenosine is an endogenous neuroprotective agent. It is of great importance to research the porcini mushrooms' adenosine for developing products. However, problems, such as the old for new and traditional methods for detecting adenosine content are complicated and time-consuming, seriously restrict industrial development. The present study aimed to achieve a rapid quantification of adenosine content in porcini mushrooms on the market using Fourier transform near-infrared (FT-NIR) spectroscopy combined with partial least squares regression (PLSR) model. Herein, the nucleoside content and spectral characteristics of the large-scale dataset (n=242) were analyzed, which was used as the calibration set for constructing the PLSR model. The PLSR model had an R2 C of 0.907 and a residual predictive deviation (RPD) of 2.726. For random samples with different origins, the R2 P was 0.768 and the RPD was 1.326, for the storage period, the R2 P was 0.952 and the RPD was 3.069, and for various collection years, the R2 P was 0.927 and the RPD was 2.548. It was demonstrated that the established method offers a rapid and reliable prediction strategy for adenosine content of random porcini mushrooms samples, which has the potential to be applied in the market.