SciLogP is a unique scientific software tool used to predict LogP values, based on 2D and 3D molecular descriptors. Included in this package is a data display for tracking trends between stereochemical properties and observed/calculated LogP values.
SciLogP Features
- LogP Predictors.
- User-defined LogP Predictor. Construct a file database, process your own training set, and use the regression analyzer to create a LogP predictor.
- Regression Analyzer. Build a multivariate regression model, the LogP predictor, using a correlation matrix to construct the best orthogonal set of descriptors.
- Interactive Graphics Window. Sort, zoom, select descriptors displayed, select axes, and display up to eight descriptors.
- Automated Processing. Construct a file database and process hundreds of compounds automatically.
- File Database. Create a compound database for LogP determinations using common, mixed 2D and 3D file formats, e.g., MOL, Hin, Al2, SKC, Smiles, or SLN.
- Output Results. Export to Microsoft® Excel in column ready format.
Advanced, neural net-based predictor for "ring" and "non-ring" compounds;
predictor based on 1400 highly diverse compounds, with validated R2 = 0.95
and test R2 = 0.94. It is a default estimator of LogP.
Five built-in linear regression expressions applicable to amides, aliphatics,
heteroaromatics, and steroids, and one all-purpose expression.