An Approach Of Biomarkers Identification On Metabolomics In Sports Training Monitoring And Its Application In Adolescent Rowing Training

Scientification is a fundamental guarantee of the effectiveness of sports training,and training monitoring is an important means of safeguarding scientific training.However,there are two major problems that need to be solved: firstly,invasive sample acquisition during training,which creates obstacles to training and competition,and affects practical operation.The second is that the specificity of training indicators is not sufficiently appreciated and there is an urgent need to establish a scientific process and method for the screening of biochemical indicators for sports training monitoring.Metabolomics technology is an emerging discipline for the systematic analysis of metabolites in samples,which can reflect changes in overall metabolic patterns in different states and screen biomarkers(Biomarker).Metabolomics research has now developed a complete process from sample collection to data analysis,however,little research has been done on the use of biomarkers for the screening of indicators for athletic surveillance.The aim of the first part is to establish a biomarker screening process for training monitoring based on metabolomics technology to provide new ideas and methods to improve the scientific,specificity and convenience of sports training monitoring;the second part investigates the effect of different intensity training on saliva metabolism by comparing different intensity training with real-world 2km training to verify the effectiveness of saliva biomarkers as training monitoring indicators.The first part establishes the biomarker screening process.A 2 min warm-up at a 50 W load was performed by using a kayak dynamometer(Dan sprint Pro,Denmark)with a260 W rated load.Blood,urine and saliva samples were collected from the athletes before and after the start of the test phase at a rated power load of 260 W.The changes in metabolic patterns induced by exercise were analysed using 1H-NMR-based metabolomics techniques by principal component analysis,(orthogonal)partial least squares(O)PLS-DA)models,and the validity of the models was tested by cross-substitution.Characteristic metabolites causing changes in metabolic patterns were identified using the OPLS-DA model with a variable importance weight value(VIP)greater than 1 as the screening criterion.The concentrations of the identified metabolites were quantified and univariate analysis(t-test,correlation analysis,etc.)was used to compare the changes in metabolites before and after exercise and to analyse the relationship between the changes in metabolites in different body fluids.operational characteristic curve(ROC)to evaluate the sensitivity and specificity of the biomarkers.The second part explores the effects of different intensities of training on salivary metabolism.Saliva samples were collected from athletes before and after the 2km live,2km test,6km test and 12 km test using the exercise protocol in Part I.The changes in salivary metabolic patterns induced by exercise were analysed by principal component analysis,(orthogonal)partial least squares(O)PLS-DA)models using metabolomics techniques with 1H-NMR and the validity of the models was tested by cross-substitution.The OPLS-DA model with a variable importance weight value(VIP)greater than 1 was used as a screening criterion to identify the characteristic metabolites causing changes in salivary metabolic patterns.The concentrations of the identified metabolites were quantified,and univariate analysis was used to compare the changes in metabolites before and after exercise,to analyse the relationship between changes in metabolites and to investigate the effects of different intensities of training on salivary metabolism,in order to verify the validity of salivary biomarkers as indicators of training monitoring.In the first part of the study results,it was found that(1)metabolomic results showed significant changes in blood,urine and saliva metabolic patterns before and after training,with the characteristic metabolites that changed in blood being pyruvate,lactate,acetylcysteine,succinate,valine,leucine and glutamine.The characteristic metabolites that changed in the urine were 3-hydroxyisovaleric acid,DMA,creatinine,glycine,hippuric acid,alanine,tyrosine,and urea.The characteristic metabolites that changed in saliva were tyrosine,formic acid,phenylalanine,methanol,acetic acid,histidine,succinic acid and urea.(2)T-test results showed that pyruvate,lactate,acetylcysteine,succinic acid and acetone passed the T-test in blood(p<0.05),while valine,leucine and glutamine did not pass the T-test.Tyrosine,formic acid and methanol in saliva all passed the T-test(P<0.05),while phenylalanine,acetic acid,histidine,succinic acid and urea did not pass the T-test.(3)Correlation and regression analyses showed that tyrosine and formic acid in saliva were significantly and positively correlated with blood lactate(r=0.863,P=0.024),and salivary formic acid was significantly and positively correlated with the amount of change in blood urea(r=-0.613,P= 0.033).(4)The results of the ROC analysis showed that the AUC for the characteristic metabolite tyrosine in saliva was 1(critical value0.0959;sensitivity 100%;specificity 100%)and the AUC for formic acid was 0.98(critical value 0.071;sensitivity 90%;specificity 100%).Tyrosine and formic acid in saliva can be used as core metabolites for exercise monitoring.In the second part of the study the 2km real-world characteristic metabolites were found to be ethanol,5-hydroxytryptophan,leucine,hydroxybutazone bodies,glutamate,tyrosine,isoleucine.2km training characteristic metabolites were tyrosine,methanol,lactate,taurine,pyruvate,formic acid.6km training characteristic metabolites were tyrosine,formic acid,ethanol,acetic acid,yohimbic acid,histidine.12 km training characteristic metabolites Urea,glutamic acid,tyrosine,ethanol.It is easy to see that the common metabolite among the characteristic metabolites for the real 2km and the different intensities is tyrosine.This suggests the validity of tyrosine as an indicator of training monitoring in junior rowing ergometers,and we can also speculate that the training load with a similar level of fatigue to the 2km real-world is 6km,rather than 2km or 12 km training.In summary we conclude that:(1)The metabolomics-based biomarker screening method has established a screening process for sports training monitoring indicators,and through correlation with classical indicators,ROC and other method validation and real-world verification,the method is scientifically feasible and project-specific,providing a new way to screen sports training monitoring indicators.(2)Salivary metabolites can be used as reliable indicators for training monitoring,providing a new idea for non-invasive and stress-free training monitoring.