Abstract
Diabetes mellitus, a chronic condition caused by defects in insulin secretion, or action, or both, is a group of metabolic disorders whose complications can contribute significantly to ill health, disability, poor quality of life and premature death. From the three main types of diabetes, Type 2 is by far the most common, accounting for about 90% of cases worldwide. Studies on the role of protein tyrosine phosphatase 1B (PTP1B) have clearly shown that it serves as a key negative regulator of insulin signaling and is involved in the insulin resistance associated with Type 2 diabetes. In the present study, a QSAR modeling work was carried out on a series of 1, 2-naphthoquinone derivatives. The inhibitory activity of such compounds was investigated by two types of QSAR methods: multiple linear regression and non-linear neural networks. This strategy afforded QSAR models with good overall accuracy and predictivity on external data, showing it to be a simple, precise and credible tool to predict and screen 1,2-naphoquinone derivatives with high inhibitory activity.
Keywords: Diabetes mellitus, Multiple linear regression (MLR), 1, 2-naphoquinone derivatives, Protein tyrosine phosphatase 1B (PTP1B), Quantitative structure-activity relationships (QSAR), Radial basis function neural networks (RBFNN), hyperglycemia, toxicology, benzothiophene, peptidomimetic derivatives
12