A Chemical Guide to Bolete Flavor

Researchers combined mass spectrometry, infrared spectroscopy, and data modeling to identify differences in mushroom flavor

October 14, 2026
3 min

The Analytical Scientist

Overview

An analytical study of five edible bolete species from southwest China identified distinct chemical profiles across 30 combined samples. The researchers measured amino acids, organic acids, and volatile compounds, and evaluated infrared spectroscopy with machine learning as a faster screening approach.

Background

Mushroom flavor reflects both nonvolatile compounds that contribute to taste and volatile compounds that produce aroma. Characterizing these compounds may help distinguish species and assess mushroom products. The study combined mass spectrometry-based chemical analysis with infrared spectroscopy and computational modeling. Its findings are limited by the number of samples and collection sites, and habitat differences may have influenced the observed profiles.

Data Highlights

Measure	Study result
Species examined	5: Boletus bainiugan, Butyriboletus roseoflavus, Rugiboletus extremiorientalis, Lanmaoa asiatica, and Phlebopus portentosus
Collection and sampling	82 fruiting bodies from southwest China; 30 combined samples, with 6 biological replicates per species
Amino acids and related metabolites	68 identified
Organic acids	57 identified
Volatile compounds	773 identified
Infrared spectroscopy and machine-learning model	Strongest model predicted selected flavor-compound concentrations with coefficients of determination of at least 0.95

Key Findings

Statistical analysis separated samples by species, indicating distinct chemical profiles among the five boletes.
Glycine, arginine, aspartic acid, lysine, and fumaric acid were among the compounds distinguishing taste profiles; sweet and umami amino acids made the largest contribution.
Six volatile compounds were selected as possible aroma markers, spanning pyrazines, an ester, an aldehyde, and a furanone associated with roasted, nutty, fruity, floral, and mushroom-like odors.
Pathway analysis linked amino-acid and organic-acid differences to possible involvement of the tricarboxylic acid and urea cycles; these relationships were inferred from chemical profiles and were not confirmed by gene-expression or enzyme-activity studies.
Computer modeling predicted that several aroma compounds could bind the odor receptor OR1A1, but the modeling did not establish how people perceive the aromas.
Attenuated total reflectance Fourier-transform infrared spectroscopy combined with machine learning predicted selected flavor-compound concentrations with a coefficient of determination of at least 0.95; the authors described it as a potential rapid initial screen, with mass spectrometry providing more detailed measurements.

Clinical Implications

The researchers described ATR-FTIR spectroscopy combined with machine learning as a potential rapid initial screen for species identification and product quality. The strongest model predicted concentrations of selected flavor compounds with coefficients of determination of at least 0.95; mass spectrometry provided more detailed chemical measurements.

Conclusion

The five bolete species had distinguishable taste- and aroma-related chemical profiles. The study reported that ATR-FTIR spectroscopy combined with machine learning predicted concentrations of selected flavor compounds with coefficients of determination of at least 0.95. Metabolic pathway and odor-receptor findings were predictions and were not experimentally confirmed.