Metabolomics Based Screening For Diagnostic Markers In Tuberculosis

Objective: 1.Review the relevant literature of TB diagnostic methods,summarize the deficiencies,and propose the necessity and significance of metabolomics studies.2.To screen for differential metabolites between individuals with active tuberculosis,latent tuberculosis infection and healthy individuals and identify potential biomarkers with diagnostic value.3.Evaluate the diagnostic efficacy of potential biomarkers and speculate on their clinical value.Methods: Metabolite detection was performed on feces from a healthy group,a latent infection group with tuberculosis,and an active tuberculosis group based on liquid-mass spectrometry(LC-MS)technology.The data were baseline filtered,peak identification,integration,retention time correction,peak to peak and normalization to obtain the final data matrix of retention time,charge to mass ratio and peak intensity,and all the charge to mass ratios could be obtained in the yin-yang ion mode.Principal component analysis(PCA),PLS-DA analysis,hierarchical cluster analysis,and KEGG enrichment analysis were used to screen the differential metabolites between the groups and draw volcano plots,heatmaps of cluster analysis,bar scatter plots,and ROC plots for visual representation of the differences.The metabolites from the screened healthy population group,latent tuberculosis infection group and active pulmonary tuberculosis group were evaluated for diagnostic efficacy and predicted their clinical diagnostic value.Results: Differential metabolites were screened between healthy,latent tuberculosis,and active tuberculosis groups.After pretreatment,a total of 1004 metabolite components were obtained in ESI mode and 702 metabolite components were obtained in ESI + mode.The results of multivariate statistical analysis showed that the model was well established and could discriminate the healthy group,latent infection group and active tuberculosis group well.Among the tuberculosis latent infection group and healthy group with VIP > 1,128 metabolites were screened in the anion mode and 95 metabolites in the cation mode.Between the active tuberculosis group and the healthy person group,334 metabolites were screened in the anion mode,and 220 metabolites were screened in the cation mode.Between the active pulmonary tuberculosis and tuberculosis latent infection groups,261 metabolites were screened in the anion mode and 176 metabolites in the cation mode.The PLS-DA results showed that the metabolites were significantly different between the active tuberculosis group,the latent tuberculosis group and the healthy population group,and the pairwise comparison between the active tuberculosis group,the latent tuberculosis group and the healthy population group showed a relatively good grouping trend either in the anionic mode or in the cationic mode.Hierarchical clustering heatmaps of pairs of group comparisons of simultaneously active tuberculosis and healthy population groups showed clear intergroup differences.Volcano plots depicted by filtering for differential metabolite significance further revealed that there were more species of differential compounds between the active tuberculosis group and the other two groups(the latent tuberculosis group and the healthy population group),while the latent tuberculosis group and the healthy population group had fewer differential compound species.Venn diagram and KEGG enrichment pathway analysis showed that three comparison pairs were co enriched for 14 compound metabolic pathways in the anionic mode and 15 compound metabolic pathways in the cationic mode.Differential metabolites were mainly enriched in drug metabolism,α-Linolenic acid metabolism,purine metabolism,and amino acid metabolism pathways.A total of 13 significant differential metabolites were identified by ROC curve analysis in pairwise comparisons among the active tuberculosis group,latent tuberculosis infection group and healthy population group,among which the differential metabolites identified in the anion mode were hypoxanthine,2 ’-deoxyinosine,tetradecan edioic acid,n-acetyl-l-ornithine,(5-l-glutamyl)-L-amino acid,13,14-dihydro-15-keto-tetranor prostaglandin D2,deoxycholic acid,2,3-Di nor-TXB2,d UMP;The differential metabolites screened in the positive ion mode were 3,4-dimethylbenzoic acid,α-Aspartylpheny lalanine,d-carnitine,N8 acetylspermidine,the above 13 compounds had AUC values > 0.70 in each comparison pair.Three potential bi omarkers,hypoxanthine,deoxycholic acid,and dump,were finally derived.Conclusion: Some changes were observed in the fecal metabolic pathways of TB patients.Combining the statistical analysis of relevant studies with this experiment,we selected hypoxanthine(hypoxanthine),deoxycholic acid(deoxycholic acid)and dump(deoxyuracil nucleotides)as potential biomarkers,among which deoxycholic acid,deoxyuracil nucleotides may only serve as potential biomarkers for the diagnosis of active pulmonary TB,while hypoxanthine may also serve as a potential biomarker for the diagnosis of latent TB infection.The above 3 potential molecular markers need further validation,and the results of this study provide an experimental basis for the screening of biomarkers in non-invasive samples for TB diagnosis.