A New Differential Gene Expression Based Simulated Annealing for Solving Gene Selection Problem: A Case Study on Eosinophilic Esophagitis and Few Other Gastro-Intestinal Diseases

Abstract

AbstractBackgroundIdentifying the set of disease-causing genes is crucial for understanding pathogenesis and developing therapies. This is particularly important to understand the pathophysiology of Eosinophilic Esophagitis (EoE) and other gastrointestinal diseases. Comparing and contrasting gene selection methods across these diseases can enhance our knowledge to identify potential therapeutic targets.MethodsThis study introduces two approaches for gene selection in gastrointestinal diseases: the Ranked Variance (RV) method and Differential Gene Expression Based Simulated Annealing (DGESA). RV acts as an initial screener by prioritizing genes based on variance. DGESA refines gene selection further by employing simulated annealing with differential expression data. We compared the outcomes of both methods through a case study on EoE and other gastrointestinal diseases.ResultsResult finds greater number of genes with negative fold changes compared to those with positive fold change in differential EoE dataset. RV Ranks top 40 genes with high variance of EoE which overlaps with the disease-causing gene set of EoE from DGESA. 40 gene pathways for each of EoE, Crohn’s Disease (CD), and Ulcerative Colitis (UC) were identified as execution outcome of our method DGESA. Among these, 10 genes for EoE, 8 for CD, and 7 for UC were confirmed in the literature for their connection with respective diseases. For EoE, 10 such confirmed genes include KRT79, CRISP2, IL36G, SPRR2B, SPRR2D and SPRR2E. For CD, the literature confirmed set encompasses NPDC1, SLC2A4RG, LGALS8, CDKN1A, XAF1, and CYBA. The validated genes in UC final gene set includes TRAF3, BAG6, CCDC80, CDC42SE2, and HSPA9.ConclusionThe RV method, serving as an initial screener, and the more refined DGESA both effectively elucidate molecular signatures in gastrointestinal diseases. Identifying and validating genes like SPRR2B, SPRR2D, SPRR2E and STAT6 for EoE showcase efficacy of DGESA. Other genes in the same pathway are interesting targets for future laboratory validation.