Gene prioritization based on biological plausibility over genome wide association studies renders new loci associated with type 2 diabetes

Abstract

Purpose: We present an approach to prioritize single nucleotide polymorphisms for further follow-up in genome-wide association studies of type 2 diabetes. Method: The proposed method combines both the use of open data access from two type 2 diabetes-genome-wide association studies (granted by the Diabetes Genetics Initiative and the Welcome Trust Case Control Consortium) and the comprehensive analysis of candidate regions generated by the freely accessible ENDEAVOUR software. Results: The algorithm prioritized all genes of the whole genome in relation to type 2 diabetes. There were six of 1096 single nucleotide polymorphisms in five genes potentially associated with type 2 diabetes: tachykinin receptor 3 (rs1384401), anaplastic lymphoma receptor tyrosine kinase (rs4319896), calcium channel, voltage-dependent, L type, alpha 1D subunit (rs12487452), FOXO1A (rs10507486 and rs7323267), and v-akt murine thymoma viral oncogene homolog 3 (rs897959). We estimated the fixed effect and P values of each single nucleotide polymorphism in the combined dataset by Mantel-Haenszel meta-analysis and we observed significant P values for all single nucleotide polymorphisms except for rs897959 at v-akt murine thymoma viral oncogene homolog 3. Conclusion: The proposed strategy may be used as an alternative tool for optimizing the information of the nearly 500,000 gene variants in which markers with modest significant P value for disease association are currently disregarded. Additionally, the said single nucleotide polymorphisms may be incorporated into the replication of the multistage design involved in the genome-wide association studies.