International Society of Bionic Engineering-Research Progress

Research Progress

MoreAnnouncement

First Announcement of the International Youth Conference of Bionic Science and …

2023-12-22

The International Youth Conference of Bionic Science and Engineering 2024 (IYCBSE 2024), hosted by the ISBE Youth Commission and jointly organized by Southeast University, Nanjing University of Aeron…

CASE collection for the ISBE website

2023-10-18

In our technical world humans continuously develop the technologies and systems to make life more comfortable and easier. Bionics, the field of translating ideas and concepts found in nature to pract…

Home > 3D Technology Platform > Research Progress > Papers > content

Research Progress

Constructing plasma protein binding model based on a combination of cluster analysis and 4D-fingerprint molecular similarity analyses

Post: 2016-03-17 02:42 View:2198

Bioorganic & Medicinal Chemistry

Volume 14, Issue 3, 1 February 2006, Pages 611–621

Jianzhong Liu^a, , ,Liu Yang^a,Yi Li^b,Dahua Pan^b,Anton J. Hopfinger^b

^a Department of Chemistry and Biochemistry, University of Delaware, Newark, DE 19716, USA

^b Laboratory of Molecular Modeling and Design (M/C 781), College of Pharmacy, The University of Illinois at Chicago, 833 South Wood Street, Chicago, IL 60612-7231, USA

Abstract

Based on 2D-connectivity molecular similarity and cluster analyses, a dataset for HSA binding is divided into the training set and the test set. 4D-fingerprint similarity measures were applied to this dataset. Four different predictive schemes (SM, SA, SR, and SC) were applied to the test set based on the similarity measures of each compound to the compounds in the training set. The first algorithmic scheme (SM), which only takes the most similar compound in the training set into consideration, predicts the binding affinity of a test compound. This scheme has relatively poor predictivity based on 4D-fingerprint similarity analyses. The other three algorithmic schemes (SM, SR, and SC), which assign a weighting coefficient to each of the top-ten most similar training set compounds, have reasonable predictivity of a test set. The algorithmic scheme which categorizes the most similar compounds into different weighted clusters predicts the test set best. The 4D-fingerprints provide 36 different individual IPE/IPE type molecular similarity measures. Further investigation shows that the NP/HA, HS/HA, and HA/HA IPE/IPE type measures predict the test set well. Moreover, these three IPE/IPE type similarity measures are very similar to one another for the particular training and test sets investigated. The 4D-fingerprints have relatively high predictivity for this particular dataset.

Keywords

Molecular similarity;4D-fingerprint similarity;HSA;Cluster analysis

Full text is available at http://www.sciencedirect.com/science/article/pii/S0968089605007935?np=y