Discovery of novel indoleamine 2, 3-dioxygenase 1 (IDO1) inhibitors by virtual screening. (February 2019)

• Record Type: Journal Article
• Authors: Zhou, Yeheng; Peng, Jiale; Li, Penghua; Du, Haibo; Li, Yaping; Li, Yingying; Zhang, Li; Sun, Wei; Liu, Xingyong; Zuo, Zhili
• Is Part Of: Computational biology and chemistry. Volume 78(2019)
• Page Start: 306
• Page End: 316
• Publication Date: 2019-02
• Subjects: IDO1 -- Molecular docking -- HipHop pharmacophore -- 3D-QSAR -- PAINS; Chemistry -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; Biochemistry -- Data processing; Biology -- Data processing; Molecular biology -- Data processing; Periodicals; Electronic journals; 542.85
• DOI: 10.1016/j.compbiolchem.2018.11.024
• Languages: English
• ISSNs: 1476-9271

Abstract

Graphical abstract: Highlights: Novel potential IDO1 inhibitors were identified by the following steps: Lipinski's Rule of Five, Veber rules, docking, pharmacophores, 3D-QSAR and PAINS filter. The best two HipHop pharmacophore models were generated based on the common features. The qualified 3D-QSAR models were obtained based on the docking conformation alignment to evaluate novel inhibitors. Six novel potential IDO1 inhibitors were identified by the combination methods. Abstract: In this study, a combination of virtual screening methods were utilized to identify novel potential indoleamine 2, 3-dioxygenase 1 (IDO1) inhibitors. A series of IDO1 potential inhibitors were identified by a combination of following steps: Lipinski's Rule of Five, Veber rules filter, molecular docking, HipHop pharmacophores, 3D-Quantitative structure activity relationship (3D-QSAR) studies and Pan-assay Interference Compounds (PAINS) filter. Three known categories of IDO1 inhibitors were used to constructed pharmacophores and 3D-QSAR models. Four point pharmacophores (RHDA) of IDO1 inhibitors were generated from the training set. The 3D-QSAR models were obtained using partial least squares (PLS) analyze based on the docking conformation alignment from the training set. The leave-one-out correlation (q 2 ) and non-cross-validated correlation coefficient (r 2 pred ) of the best CoMFA model were 0.601 and 0.546, and the ones from the best CoMSIA model were 0.506 and 0.541, respectively. Six hits from Specs database were identified and analyzed to confirm their binding modes and key interactions to the amino acid residues in the protein. This work may provide novel backbones for new generation of inhibitors of IDO1.