PhysChem Suite

ACD/PhysChem Suite is a collection of prediction modules from ACD/Labs designed to calculate physicochemical properties from molecular structure. It supports structure-based predictions of properties including logP, logD, pKa, aqueous solubility, boiling point, vapor pressure, Sigma, and other molecular descriptors for organic compounds. The tool is intended for medicinal chemists, synthetic chemists, and research scientists involved in QSPR screening, high-throughput library evaluation, and lead optimization.

Calculations can be performed on single compounds or libraries of tens of thousands of structures. Models for logP, pKa, logD, and aqueous solubility are trainable with experimental data, and the platform supports inclusion of custom in-house models and third-party prediction algorithms.

Core Capabilities

• Predict aqueous solubility, boiling point, logP, logD, pKa, Sigma, and additional molecular descriptors for organic compounds up to a recommended molecular weight of ≤2000 Daltons, including peptides, proteins, polymeric units, and other beyond rule-of-5 (bRo5) compounds
• Input structures by drawing in-app, copy/paste from third-party drawing packages, SMILES string, InChI code, imported MOL/SK2/SKC/CDX files, or by name search using a built-in dictionary
• Multiple algorithms available for many properties, with automatic detection of tautomeric forms where applicable
• Calculate properties for compound libraries with tools to sort, filter, plot, and rank results; set user-defined label colors and filter results numerically
• View activity history for previously calculated values
• Report results to PDF or copy to external applications; download QPRF and QMRF documents for ACD/LogP (GALAS model) and ACD/LogS0
• Add custom models or in-house prediction algorithms via web service connection using an XML protocol or as a DLL

Reliability and Result Evaluation

• Assess prediction reliability using 95% confidence intervals, a Reliability Index, and display of the five most similar structures from the training library with experimental values and literature references
• View interactive calculation protocols that detail contributing functional groups, atoms, and interactions
• Color-map substructure and atomic contributions to property values directly on the structure

Property-Based Structure Modification and Lead Optimization

• Investigate structure-property relationships and understand the pharmacokinetic profile of lead compounds
• Evaluate good/bad indicators for Lipinski's rule-of-5 and lead-likeness
• Identify structural fragments associated with toxicity
• Generate libraries of analogs with substituent modifications targeting an optimal property profile using the interactive optimization tool
• Sort, filter, and prioritize hundreds of structural analogs according to a desired property profile
• Create and use custom fragment libraries; target synthetically accessible fragments with a built-in retrosynthesis tool

Aqueous Solubility Module

• Calculate quantitative solubility in pure (unbuffered) water at 25°C
• Predict qualitative solubility at pH 7.4, categorizing compounds from highly soluble to insoluble using the GALAS algorithm
• Estimate intrinsic solubility (logS0) based on a training set of over 6,800 compounds using the GALAS algorithm, with reliability values and up to five similar structures from the training set
• Predict pH-dependent aqueous solubility (logS) at physiologically relevant pH values (1.7, 4.6, 6.5, 7.4, 8.0) with a plot of predicted pH versus solubility
• Train the model with experimental values

Boiling Point and Vapor Pressure Module

• Estimate boiling point of organic compounds as a function of pressure
• Predict vapor pressure as a function of temperature
• Estimate enthalpy of vaporization at the boiling point
• Estimate flash point in the temperature unit of choice
• View results in table or graphical plot format

LogP Module

• Predict logP using three algorithms: Classic, GALAS, and a Consensus model based on the other two
• Detailed calculation protocol listing all contributing functional groups, carbon atoms, and interactions through aliphatic, aromatic, and vinylic systems (Classic algorithm)
• Color highlighting of hydrophilic and lipophilic substructures on the molecule (GALAS algorithm)
• Train the model with experimental values to improve predictions for proprietary chemical space; create and select different training libraries or revert to the built-in algorithm

LogD Module

• LogD predictions are derived from the logP and pKa models within PhysChem Suite
• Select from available logP and pKa algorithms (default: logP Consensus, pKa Classic)
• View logD results by pH, including physiologically relevant values (1.7, 4.6, 6.5, 7.4, 8.0); interact with the logD vs. pH plot to retrieve values at any pH of interest
• Train the model with experimental logP and pKa values; create and manage separate training libraries

pKa Module

• Calculate the acid dissociation constant (pKa) under standard conditions (25°C, zero ionic strength) in aqueous solution for every ionizable group
• Choose from two algorithms: ACD/pKa Classic (default) and GALAS
• Color-coding of ionizable groups: red for acidic, blue for basic, purple for amphoteric centers; color intensity indicates acid/base strength
• Reliability range in ±log units for calculated pKa values
• Detailed calculation protocol for each ionization stage with interactive structure highlighting
• Train the algorithm with experimental data
• GALAS model features include: percentage contribution of individual ionization microstages; interactive plots and tables of pKa values as a function of pH (0–14); net charge vs. pH; protonation state vs. pH; ionogenic group state vs. pH; and fraction of all ionic forms at a selected pH

Sigma Module

• Calculate substituent-specific parameters for selected molecular fragments in aqueous solution at zero ionic strength and 25°C
• Parameters include the electronic substituent constant (Hammett), steric constants (molar volume, molar refractivity), and the hydrophobic constant (Hansch Pi)

Workflow

1. Draw or import a structure
2. Select the property of interest
3. Review results and make decisions
4. Report to PDF or copy/paste to another application

PhysChem Suite is available as part of the ACD/Labs Percepta platform, which supports enterprise deployment. The platform has been deployed at research facilities including Merck Germany, where it is used to support medicinal chemists in synthesis planning and optimization of new chemical entities. The tool's configurability and ability to incorporate in-house algorithms were cited as key factors in enterprise adoption.