Serum FCN-3 and TMOD-4 Can Predict Motion Sickness by Proteomics and Machine Learning
LIU Zhi
ZHANG Jinhong
ZHANG Chun
WANG Xiaoyu
LI Jinrang
Abstract：Objective To identify to explore proteins associated with motion sickness(MS)by proteomics techniques and machine learning.Methods Serum samples were collected from 51 patients with MS and 68 controls.Differentially expressed proteins(DEPs)were analyzed and discovered using proteomics techniques and bioinformatics methods,as well as predictive modeling using machine learning.Candidate proteins were validated by ELISA in a separate cohort.Results A total of 27 DEPs(11 up-regulated and 16 down-regulated)were identified to differ significantly between MS patients and controls.Functional enrichment analysis showed that DEPs were mainly enriched in platelet activation,ions binding,cellular exosomes,neurodegeneration,amyotrophic lateral sclerosis,Huntington's disease,immunity,and hemostasis.Based on multiple machine learning and ROC curve analyses,we constructed a potential diagnostic model based on the levels of the best 8 DEPs in serum with 95.8%specificity and 100%sensitivity(AUC=0.997,P<0.001).Among the eight DEPs,FCN-3 and TMOD-4,the two most abundant proteins,were selected as another potential model with 75.0%specificity and 91.7%sensitivity(AUC=0.870,P<0.001).ELISA results were consistent with proteomics analysis.Conclusion Our preliminary data suggest that MS and non-MS patients have different serum protein expression profiles.In addition,alterations in FCN-3 and TMOD-4 may serve as novel biomarkers for the diagnosis of MS.
Keywords：Motion sicknessproteomicsFCN-3TMOD-4machine learning
Publication Date：2025-08-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：8( 669-676 )