An overall workflow based on gas chromatography-mass spectrometry (GC-MS) was established for the analysis of the serum amino acid profile between urolithiasis patients (n=80, age (46.82±13.39) years) and healthy controls (n=37, age (43.46±12.79) years). The raw data from GC-MS analysis were processed by multivariate statistical methods to build the model. Following this, student's t-test and logistic regression were performed and receiver operator characteristic (ROC) curve was plotted to identify the potential biomarkers. Good linearities were observed for the target amino acids, with correlation coefficients (R2) greater than 0.9985. The limits of detection (LODs) were 0.1-4.0 μmol/L. The results indicated a significant discrimination between the urolithiasis and control groups. Five significantly differentially expressed amino acids (variable importance in projection (VIP)>1 and p<0.05) were found to provide the scientific evidence for the early diagnosis of urolithiasis, while the sensitivity of the integrated five differential amino acids was up to 97.3%. In particular, the area under the curve (AUC) of serine reached 0.819, which suggested a great clinical screening value.
Keywords: amino acid profile; gas chromatography-mass spectrometry (GC-MS); potential biomarkers; urolithiasis.