Recommendations for CoMFA Studies and 3D QSAR Publications

Abstract

2 3Authors of CoMFA publications should consider the most important facts and prerequisites of 3D QSAR analyses and should publish a minimal set of information on their results, following these recommendations [I, 21: Selection of starting geometries.Especially in the case of flexible molecules authors should rationalize their choice by describing the approach used. In general, conformations of flexible drug molecules, adopted e.g. in vacuum, in solution, or in the crystal, are not necessarily similar to those required at a protein binding site. The conformations should be selected in a way to represent -to the best knowledge -the conditions under which the system was studied. In any case, the authors should give some ideas why the molecules in the data set were selected in these particular conformations [3 -61. Methods of geometry optimization of 3D structures.Starting structures obtained from model building or from any other source, including experimentally determined structures (e.g. X-ray structures, structures derived from NOE information, see section I), are in most cases optimized by some kind of molecular mechanics or quantum chemical approach. A documentation of the method (e.g. MM2, MAXIMIN, etc.) and its parametrization is required. It should be described whether the optimization has been performed including or ignoring Coulombic terms. If atom types not present in a standard force field implementation have been considered it should be mentioned how they were incorporated into the optimization.For quantum chemical optimization techniques, the authors should describe which methods have been applied (e.g. semiempirical methods such as CNDO, MNDO, AMI, PM3, PCILO, etc., or ab initio methods) and which convergence criteria have been set. For geometries obtained from ab initio calculations, the basis set should be given.Information about the quality of the structural data is important. Due to the functional forms of the fields, the CoMFA approach is very sensitive to small variations in geometry. If, e.g. the convergence criteria are too loosely set, different geometries will be obtained for identical portions of structural analogs, only because the next local minimum has not been reached. This will result in small RMS deviations which cause corresponding variances in the CoMFA fields [5, 61. 1 Charges used in CoMFA.A list of atomic charges is required for the calculation of electrostatic fields within CoMFA. The technique used in the analysis should be given together with the formal overall charge of the molecule (e.g. the protonation state). Some proof of consistency among the different methods is highly recommended IS]. Description of alignment criteria, grid spacing, and size of the CoMFA lattice and probe atodprobe charge used.Information about the alignment of the molecules is one of the most important criteria to decide on the quality of a 3D QSAR study. Accordingly, it should be clearly stated whether the alignment is based on experimental evidence or on hypothetical assumptions. Any computational ...

Structure‐Lipophilicity and Structure‐Polarity relationships of amino acids and peptides

“…In all equations, the 95 % confidence intervals, estimated with TSAR using the jackknife approach, are given in parenthesis. Following recent rules [18] [19], the crossvalidated r-square (q2) is also given, together with the conventional r-square (r*), the standard error (s), and the F value (F). In the discussion, all statistical models with a standard error s inferior or equal to the uncertainty of measurements were considered as significant.…”

Diosgenone Synthesis, Anti-Malarial Activity and QSAR of Analogues of This Natural Product

“…This field gave information concerning the model's salient features, specified by QSAR coefficients' magnitude, scaled by associated variability at each grid point, specified by standard deviation [12]. Regarding steric contribution, increased steric volume beyond the D-ring (green mesh) correlated with positive contribution to bioactivity, whereas the yellow mesh around the A-ring meant negative steric field contribution to bioactivity.…”

“…A final model was obtained by non-cross-validated PLS analysis for ascertaining the optimum number of components. The experimental activity values were transformed to a logarithmic scale to improve data distribution, as recommended by usual QSAR practice [12]; the transformation used was thus Log (1/IC 50 ).…”

“…This involved using 3D quantitative structure activity relationship (3D-QSAR) [12], a well-established drug design methodology providing information regarding statistical correlations between molecular structure descriptors and experimentally-measured bioactivity [13]. The partial linear squares (PLS) [14] algorithm provided a solid mathematical background for 3D-QSAR for statistically-significant models having acceptable predictive capability, ensured by very conservative cross-validation testing.…”

Integrated Approach Using Protein and Ligand Information to Analyze Selectivity‐ and Affinity‐Determining Features of Carbonic Anhydrase Isozymes

“…Therefore, we tried to profit in the present study from an enlarged data basis and compiled a better distributed and balanced data set. For QSAR studies, the affinity data should spread over a range of at least three logarithmic units [41] and display approximately an even distribution. In the present study, these requirements are particularly difficult to fulfill, since we have to deal with six dependent variables simultaneously (three pK i values for the three isozymes and three pairwise differences of the pK i values (vide infra)).…”