A proof-of-concept application of gene set enrichment analysis principles to olfactory phenotyping in Parkinson's disease.

Introduction: While gene set enrichment analysis (GSEA) framework is widely used on transcriptomics or proteomics data, its applications beyond "omics" data has not been extensively explored. Its principal nature enables it to apply to any dataset where variables can be rationally sorted into categories, or sets. This study demonstrates a novel application of GSEA outside of its traditional use by applying the framework to clinical olfactory phenotyping data. Methods: We applied the GSEA framework to University of Pennsylvania Smell Identification Test (UPSIT) results from Parkinson's disease patients and healthy controls. Odors were classified into seven distinct functional categories, and a rank-based running sum was used to identify coordinated changes across those categories. Results: This approach revealed an exploratory pattern of negative enrichment within the "citrus/fresh" odor category in Parkinson's disease patients. The computational pipeline was systematically evaluated alongside synthetic positive, negative, and permutation-based sanity checks. Discussion: Category-specific patterns of olfactory loss may offer additional insight into neurodegenerative processes, suggesting potential value for more refined olfactory phenotyping in Parkinson's disease. This proof-of-concept highlights how established genomic algorithms can be directly adapted to clinical tools, questionnaires, or cognitive data where variables can be meaningfully grouped. However, the results here should be considered exploratory.