Evidence-Based Prioritization and Scoring of Genomic Loci Interacting with Pioglitazone Therapeutic Response in Diabetic Patients.

Abstract

Extensive literature shows multiple genetic loci's involvement in regulating pioglitazone's action in diabetic patients; most of the available literature lacks replication, and it is difficult to separate true positives from false-positive loci. To overcome the inconsistency in pharmacogenetics and pharmacogenomics data, we performed evidence-based semi-automated prioritization and scoring of candidate genes of pioglitazone in diabetic patients. We developed a Python script and searched with MeSH terms of "Pioglitazone (PIO)" and synonyms," and (i) built a complete library of published data (22,532 publications), (ii) extracted sentences containing pioglitazone-gene pair by using Python 3.6, (iii) annotated each sentence for its relevance, (iv) developed evidence scoring algorithm based on pharmacogenomics relatedness, frequency, and consistency. From the literature, we have identified that the gene encoding peroxisome proliferator-activated receptor gamma showed the strongest evidence of mediating the pioglitazone response. Additionally, polymorphisms in genes affecting the pharmacokinetics were from the Cytochrome enzyme family, i.e., Cyp2C8, Cyp3A4, and Cyp2C9, and pharmacodynamics were protein tyrosine phosphokinase (Ptprd), Lipoprotein (Lpl), Adiponectin receptor 2 (Adipora2), and PPARG coactivator 1 alpha (Ppargc1). Furthermore, drug-induced adverse drug reactions were associated with polymorphisms in the Solute carrier family 12 member 1 (Scl12a1) and Aquaporin 2 (Aqp2) genes. We have prioritized candidate gene variants and their response to pioglitazone in diabetic patients. Our results showed PPAR-α receptors/ PPARG as one of the most robust evidence of therapeutic response to pioglitazone and highlighted other variants in different pathways. However, we further warrant validation of our results in a large drug response GWAS dataset.