Increase in number of people diagnosed with diabetes makes this disease a new health threat in the 21st century. Understanding the etiology of and finding a way to prevent diabetes, especially type 2 diabetes mellitus, is an urgent challenge for the health care community and our society. Pancreatic islet cells are responsible for maintaining normal blood glucose level and if there is any disturbance that leads to the onset of diabetes. Human pancreatic islet cells contain α,β,δ, and PP cells. Understanding the contribution of each type of cell through gene expression in type 2 diabetes mellitus is very important for the development of diagnostic tools. Therefore, gene expression data of α,β,δ and PP cells can be used. Single cell RNA sequencing technology has been found useful to generate expression data for individual cells. The gene expression data is usually used to find genes that are related to clinical outcome. However, in a biological process a set of genes are involved that share functional similarity. Analysing only single type of data may not generate significant type 2 diabetes mellitus genes. In this regard, an integrated approach has been used to analyse single-cell RNA sequencing data of human pancreatic islet cells. The integrated approach is designed by incorporating protein-protein interaction network data and gene expression data to select a set of genes that are highly related to diabetes also they are functionally related among themselves. The effectiveness of the approach is demonstrated over other existing methods. © 2018 IEEE.