Clustering Analysis Indicates Genes Involved in Progesterone-Induced Oxidative Stress in Pancreatic Beta Cells: Insights to Understanding Gestational Diabetes

Clustering analysis in gene expression data has been shown to be useful for understanding gene function, gene regulation, and cell processes and subtypes. Due to the wide availability of techniques for this task, the choice of an appropriate method is critical. Trying to mitigate this problem, Saelens and coauthors performed, in 2018, a benchmark study based on external validation indices. The present work proposes an extension of this analysis by including internal indices and applying it in a study case to investigate gestational diabetes through experiments on microarray data of pancreatic beta cells submitted to supra-pharmacological doses of progesterone. The results of the clustering method selected by the proposed extension have shown to be helpful in an enrichment analysis that identified TXNIP gene as relevant for future work aiming at understanding in more details the gestational diabetes phenomena.