LogP

The LogP software predicts the octanol-water partition coefficient (logP), a measure of hydrophobicity, from chemical structures. It is designed to calculate logP for neutral molecules, identify hydrophilic and hydrophobic fragments, and train algorithms with experimental data.

LogP offers three prediction algorithms: Classic, GALAS, and Consensus. The Classic algorithm uses over 12,000 experimental values, focusing on isolating carbons. GALAS adjusts predictions based on similarity to a training set of over 22,000 compounds. The Consensus model combines both algorithms, weighting them according to their reliability in specific chemical spaces.

Key features include the ability to draw or import structures, use SMILES strings, InChI codes, and various file formats. Users can create scatter plots, filter, sort, rank, and prioritize compound libraries. The software also provides 95% confidence intervals for logP values and highlights hydrophilic/hydrophobic substructures.

• Deeper Insights: Easily create scatter plots and manage compound libraries.
• Three-in-One Calculators: Access results from Classic, GALAS, and Consensus algorithms.
• Customizable: Train the model with in-house data for improved accuracy.

LogP can be deployed as a desktop application, batch processing tool, or browser-based application, compatible with Windows and Linux. It integrates with corporate intranets and workflow tools like Pipeline Pilot and KNIME.

The software is beneficial across various industries, including pharmaceuticals, agrochemicals, environmental science, and consumer products, aiding in drug development, pollution assessment, and product formulation.

Recent updates in LogP v2025 include an expanded GALAS training set and improved coverage of complex chemical spaces, enhancing prediction accuracy.