Urine and serum metabolic profiling combined with machine learning for autoimmune disease discrimination and classification.
Qiuyao DuXiao WangJunyu ChenCaiqiao XiongWenlan LiuJianfeng LiuHuihui LiuLixia JiangZongxiu NiePublished in: Chemical communications (Cambridge, England) (2023)
Precision diagnosis and classification of autoimmune diseases (ADs) is challenging due to the obscure symptoms and pathological causes. Biofluid metabolic analysis has the potential for disease screening, in which high throughput, rapid analysis and minimum sample consumption must be addressed. Herein, we performed metabolomic profiling by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) in urine and serum samples. Combined with machine learning (ML), metabolomic patterns from urine achieved the discrimination and classification of ADs with high accuracy. Furthermore, metabolic disturbances among different ADs were also investigated, and provided information of etiology. These results demonstrated that urine metabolic patterns based on MALDI-MS and ML manifest substantial potential in precision medicine.
Keyphrases
- machine learning
- mass spectrometry
- deep learning
- liquid chromatography
- high throughput
- multiple sclerosis
- artificial intelligence
- single cell
- big data
- high performance liquid chromatography
- capillary electrophoresis
- high resolution
- ms ms
- risk assessment
- healthcare
- depressive symptoms
- drug induced
- social media
- atomic force microscopy
- simultaneous determination
- solid phase extraction