Extended data for "Pre-diagnostic Changes in the Metabolome of Inflammatory Bowel Disease"

Extended data for "Pre-diagnostic Changes in the Metabolome of Inflammatory Bowel Disease"Extended Data Table 1. Annotation of metabolite features. 2D structure annotations are considered to be at a confidence level of 2 and class annotations at level 3. GNPS and CSI+FingerID were used to achieve structural annotations. Annotations were curated manually to improve consistency of nomenclature and to evaluate cases when annotations from CSI+FingerID and GNPS were not in agreement. Cosine scores for GNPS are presented as gnps_score and only considered at cosine scores >0.7. CSI+FingerID scores are used to evaluate the best match for each Metabolite ID but are less informative in evaluating the annotation confidence between different Metabolite IDs. For each Metabolite ID, database links to the best matching metabolite (CSI+FingerID) are found under csi_links. All class annotations were determined by CANOPUS, and probabilities at different metabolite class levels (according to Classyfire) are presented. The class given under most_specific_class was determined by CANOPUS and the corresponding most_specific_class_probability refers to the predicted probability that the Metabolite ID is a metabolite belonging to the given class at that level. Extended Data Table 2. Associations between metabolites and future development of inflammatory bowel disease, Crohn’s disease and ulcerative colitis. All results are calculated using conditional logistic regression models. p= p-value, fdr=false discovery rate adjusted p-value, OR = odds ratio, lowint = lower boundary of 95% confidence interval and highint = upper bound of 95% confidence interval. The associations were considered significant at fdrExtended Data Table 3. Results from metabolite enrichment analyses for Crohn’s disease. P-values are calculated using Fisher’s exact test. False discovery rate adjusted p-values are presented as fdr and considered significant at fdrExtended Data Table 4. Metabolite selections for prediction model of Crohn’s disease presented for each iteration of the model. Extended Data Table 5. Metabolite selections for prediction model of ulcerative colitis presented for each iteration of the model. Extended Data Table 6. Metabolite selections for prediction model of inflammatory bowel disease presented for each iteration of the model.