Journal Article

Analysis of Changes in Taste Characteristics of Coffee at Different Primary Processing Methods Using E-Tongue, Untargeted Metabolomics and WGCNA

Yuan, Yaqian, Chen, Chun, Fu, Xiong, Chen, Weijun, He, Rongrong, Chen, Wenxue, Chen, Haiming, Wang, Jia, Pei, Jianfei, Liang, Ying, Zhong, Qiuping

2026

Overview

The primary processing shapes the taste characteristics of coffee beans, while the regulation pathways remain unclear. Coffee beans processed by five methods-dry processing (DP), wet processing (WP), red honey (RH), black honey (BH) and anaerobic fermentation (AF)-were evaluated using electronic tongue analysis, sensory evaluation, and untargeted metabolomics. Sensory evaluation scores for mouthfeel, balance, and overall were higher in BH and AF. Conversely, the WP and DP exhibited heightened bitterness and astringency responses on the electronic tongue sensors, particularly for the former. The multigroup metabolomic comparison identified 808 DMs, and WGCNA revealed eight sensory-related modules containing 467 hub metabolites, mainly amino acids and derivatives, organic acids, alkaloids, and phenolic acids. KEGG analysis demonstrated that pathways such as caffeine metabolism and glycerophospholipid metabolism were the main pathways responsible for the metabolic differences. Further correlation analysis revealed potential flavor components closely associated with key taste characteristics. 1,3,4,5-tetrahydroxycyclohexanecarboxylic acid and Tyr demonstrated positive associations with bitterness, while TPC, TFC, Gly, and Met exhibited negative correlations with bitterness and astringency. Glu demonstrated a positive correlation with umami. These findings elucidate the material basis by which the primary processing modulates non-volatile compounds and taste perception, offering new insights into enhancing coffee quality.

• Publisher: MDPI AG,Multidisciplinary Digital Publishing Institute (MDPI)
• Subject: Acids / Aluminum / Amino acids / Anaerobic processes / Astringents / Bitterness / Caffeine / Coffee / Coffee industry / Correlation analysis / Electronic tongues / Fermentation / Glycine / Hainan Robusta coffee / Honey / Humidity / Information processing / Metabolites / Metabolomics / Methods / Organic acids / Phenolic acids / primary processing methods / Quality standards / Sensors / Sensory evaluation / Sodium / Taste / taste characteristics / Taste perception / Umami / untargeted metabolomics / Volatile compounds / weighted gene co-expression network analysis / Wet processing