Internet Electronic Journal of Molecular Design - IEJMD, ISSN 1538-6414, CODEN IEJMAT
ABSTRACT - Internet Electron. J. Mol. Des. August 2007, Volume 6, Number 8, 218-228 |
Topological Descriptors in Modeling Tumor Necrosis Factor alpha Inhibitory
Activity of Xanthines, Pteridinediones and Related Compounds
Hemant Kumar Jain and Ram Kishore Agrawal
Internet Electron. J. Mol. Des. 2007, 6, 218-228
|
Abstract:
Inhibition of tumor necrosis factor alpha (TNF-α) is an important strategy for the
treatment of various inflammatory, infectious, immunological, or malignant
diseases such as Crohn's disease, rheumatoid arthritis and psoriasis. Therefore,
the TNF-α inhibitory activity of xanthines, pteridinediones and related
compounds has been analyzed with different topostructural and topochemical
indices. The quantitative structure-activity relationships (QSAR) study was
performed using derivatives of xanthines, pteridinediones and related compounds
with Dragon 3.0 structural descriptors. The relationship between inhibitory
activity and various descriptors is established by step-wise multiple regression
analysis using Systat 10.2 and Valstat. The analyses have generated significant
and predictive QSAR models. The values of statistical indices for QSAR with
simple topological descriptors is R = 0.91, F = 19.92,
SEE = 24 and R2CV = 0.68;
for QSAR with more complex topological descriptors is R = 0.90, F = 25.74,
SEE = 0.23 and R2CV = 0.67
and for QSAR with mixed classes of descriptors is
R = 0.93, F = 25.93, SEE = 0.21 and
R2CV = 0.73. These models suggest that Lopping
centric index, fourth order bond information content of neighbourhood symmetry,
polarizability and electronegativity weighted Geary graph spatial autocorrelation
coefficient of the first lag, atomic masses weighted Moran graph spatial
autocorrelation coefficient of the third lag, and average connectivity index χ1 have
a direct correlation with the biological activity. The average valence connectivity
index χ3, electronegativity weighted Geary graph spatial autocorrelation
coefficient of the seventh lag, electronegativity weighted Moran graph spatial
autocorrelation coefficient of the fifth lag and number of substituted aromatic
sp2 carbon have an inverse correlation with the biological activity.
|