QSAR Study on tadpole narcosis

Diagnostic Tools to Determine the Quality of “Transparent” Regression-Based QSARs: The “Modelling Power” Plot

“…Toxicity Data. The concentrations of 123 compounds causing 50% lethality ( C nar , in mol dm -3 ) to the tadpole were taken from ref . It should be noted that the data set presented by Agrawal et al contains duplicate entries (without comment) of six compounds (butan-1-ol, ethanol, methanol, N -ethylurethane, octan-1-ol, and propan-1-ol).…”

“…This is to simplify the tasks related to the comparison of models (i.e., variable selection, validation schemes, impact of outliers, number of compounds and data range, MLR vs PLS algorithms, etc.). To investigate the use of the modeling power plot, one of the oldest and most fundamental toxicological databases has been investigated, the narcotic activity to tadpoles [compiled from original data by Abraham et al and analyzed recently by Agrawal et al] . The advantage of this approach is that it forms a historical and well-recognized database, and the mechanism of toxic action (narcosis) is fundamental and its modeling is well understood.…”

Application of the modelling power approach to variable subset selection for GA-PLS QSAR models

“…The concentrations causing 50% lethality (C nar , in mol dm −3 ) of 123 compounds to the tadpole were taken from [12] (the full data matrix is available from this reference). The response variable (y-vector) was taken as log 1/C nar [6].…”

“…All of the compounds are assumed to be narcotic in action and cover a broad range of organic structures [12]. In a previous work [6], it has been show that four compounds (Triacetin, Hexan-1-ol, Decan-1-ol and Acetal) appear as outliers.…”

“…See ref. [12] and references therein for more information on the descriptors. The 10 predictor variables were used as initial X-matrix (X 119 × 10 ).…”