Integrating Multi‐Omic Data from Proteome, Phospholipidome, and Oxylipid Profiles Explains a Sizable Proportion of Variance Associated with the Presence of Colon Polyps

Changes in the proteome and lipidome are associated with increased inflammation, inflammation‐induced colon carcinomas, and colon tumor survival. Our lab has previously reported levels of specific serum proteins and plasma fatty acids are associated with colon polyps. The objective of this study was to determine associations with colon polyps and the proportion of variance explained by integrating entire proteomic and lipidomic profiles. A total of 122 Caucasian males (48–65 years) were categorized by colon polyp type, either: no polyps, 1 or more hyperplastic polyps, or 1 or more adenomas. ELISA was used to quantify 70 serum proteins. The plasma phospholipidome (>2000 lipids) was profiled using UPLC‐untargeted high‐resolution mass spectrometry. Plasma non‐esterified fatty acids and >60 oxylipids were fractionated using solid phase extraction and quantified using HPLC‐MS/MS. All ‐omic data sets were centered and scaled in the same manner. Bayesian generalized additive models were used to integrate the high‐dimensional multi‐omic profiles. Single marker regressions were employed to assess each lipid with polyp type. All models were adjusted for age and smoking, and p‐values were adjusted for false discovery. The serum proteome explained approximately 14% of the variation in polyp type. Of the >2000 lipids in the plasma phospholipidome, none were individually associated with polyp type after false discovery correction. However, the entire plasma phospholipidome accounted for over 20% of the variation in polyp type, which indicates a general change of the lipid profile in different polyp types, rather than a specific change in levels of a specific lipid. In plasma oxylipid single marker regressions, arachidonic acid‐derived oxylipids were significantly elevated in those with polyps. Entire oxylipid profiles only accounted for a marginal (< 9%) amount of variation associated with polyps. When integrating the multi‐omic profiles, there was similar variation explained by the plasma phospholipidome whether adding the serum proteome or plasma oxylipids to the model. Integrating the three ‐omic data sets revealed a decreased amount of variation that each ‐omic profile could explain for polyp type, compared to integrating only 2 of ‐omic data sets. This suggests that changes in biological markers across ‐omics may be correlated. In conclusion, the entire plasma phospholipidome accounted for the largest proportion of variation associated with polyp type, compared to our serum protein and plasma oxylipid profiles. Future studies should investigate the predictive ability of the phospholipidome in larger more diverse populations, and determine if integrating the lipidome with other ‐omic data (i.e., genome, methylome, or gene expression) improves the variation explained with polyp type. Support or Funding Information USDA NIFA 201‐67011‐25205