Premium
coral Software: QSAR for Anticancer Agents
Author(s) -
Benfenati Emilio,
Toropov Andrey A.,
Toropova Alla P.,
Manganaro Alberto,
Gonella Diaza Rodolfo
Publication year - 2011
Publication title -
chemical biology and drug design
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.59
H-Index - 77
eISSN - 1747-0285
pISSN - 1747-0277
DOI - 10.1111/j.1747-0285.2011.01117.x
Subject(s) - quantitative structure–activity relationship , set (abstract data type) , predictability , molecular descriptor , computer science , calibration , coral , training set , artificial intelligence , property (philosophy) , data set , data mining , biological system , machine learning , mathematics , statistics , biology , ecology , programming language , philosophy , epistemology
CORrelations And Logic (coral at http://www.insilico.eu/coral) is freeware aimed at establishing a quantitative structure - property/activity relationships (QSPR/QSAR). Simplified molecular input line entry system (SMILES) is used to represent the molecular structure. In fact, symbols in SMILES nomenclatures are indicators of the presence of defined molecular fragments. By means of the calculation with Monte Carlo optimization of the so called correlation weights (contributions) for the above-mentioned molecular fragments, one can define optimal SMILES-based descriptors, which are correlated with an endpoint for the training set. The predictability of these descriptors for an external validation set can be estimated. A collection of SMILES-based models of anticancer activity of 1,4-dihydro-4-oxo-1-(2-thiazolyl)-1,8-naphthyridines for different splits into training and validation set which are calculated with the coral are examined and discussed. Good performance has been obtained for three splits: the r(2) ranged between 0.778 and 0.829 for the sub-training set, between 0.828 and 0.933 for the calibration set, and between 0.807 and 0.931 for the validation set.