Replacement Method and Enhanced Replacement Method Versus the Genetic Algorithm Approach for the Selection of Molecular Descriptors in QSPR/QSAR Theories

ANDREW G. MERCADER; PABLO R. DUCHOWICZ; FRANCISCO M. FERNANDEZ; EDUARDO A. CASTRO

JOURNAL OF CHEMICAL INFORMATION AND MODELING

AMER CHEMICAL SOC

Año: 2010 vol. 50 p. 1542 - 1548

1549-9596

We compare three methods for the selection of optimal subsets of molecular descriptors from a much greater pool of such regression variables. On the one hand is our enhanced replacement method (ERM) and on the other is the simpler replacement method (RM) and the genetic algorithm (GA). These methods avoid the impracticable full search for optimal variables in large sets of molecular descriptors. Present results for 10 different experimental databases suggest that the ERM is clearly preferable to the GA that is slightly better than the RM. However, the latter approach requires the smallest amount of linear regressions and, consequently, the lowest computation time.