ACD Labs Logo

PhysChem, ADME & Toxicity Calculations with Percepta Software

Predict Molecular Properties—Physicochemical, ADME & Tox—from Chemical Structure

  • Predict molecular properties from structure, name, or SMILES string (pKa module)
    Predict molecular properties from structure, name, or SMILES string
  • Percepta offers a single interface for a complete portfolio of property predictions
    Percepta offers a single interface for a complete portfolio of property predictions
  • Sub-structure highlighting provides a visual aid to understanding prediction results (LogP Module)
    Sub-structure highlighting provides a visual aid to understanding prediction results
  • Plot, sort, and rank results with ease (Percepta Portal—web application)
    Plot, sort, and rank results with ease

Industry leading calculators for physicochemical properties, ADME properties, and toxicity endpoints.

Molecular Property Calculations on the Percepta Platform

Percepta offers calculation of a variety of physicochemical (PhysChem) properties, ADME properties, and toxicity endpoints. See the full list of molecular property calculators, with details, below:


  • Aqueous Solubility

    Aqueous (Water) Solubility Module

    • Calculates pH dependent aqueous solubility, intrinsic solubility, and solubility of the chemical dissolved in pure (unbuffered) water at 25°C and zero ionic strength; along with the equilibrium pH of the solution
    • The model is trainable with experimental values to improve predictions for proprietary chemical space
  • Boiling Point/Vapor Pressure

    Boiling Point/Vapor Pressure Module

    • Estimates the boiling point of organic compounds at specified pressure
    • Predicts the vapor pressure of organic compounds as a function of temperature
  • LogD

    LogD Module

    Calculate logD (lipophilicity based on pH for ionizable compounds) from structure.

    Learn more

  • LogP

    LogP Module

    Calculate logP (lipophilicity) from structure.

    Learn more

  • pKa

    pKa Module

    Calulate pKa values (acid dissociation) from structure.

    Learn more

  • Sigma

    Sigma Module

    • A model for calculation of substituent-specific parameters for selected fragments of the molecule:
      • Eectronic constants (Hammett Sigmas)
      • Steric constants (molar volume, molar refractivity)
      • Hydrophobic constant (Hansch Pi)
  • Other PhysChem Descriptors

    Other PhysChem Descriptors

    Molecular property calculators are also available for the following:

    • Density
    • Freely Rotatable Bonds
    • H-Bond Donors and Acceptors
    • Index of Refraction
    • Molar Refractivity
    • Molar Volume
    • Molecular Weight
    • Parachor
    • Polar Surface Area
    • Polarizability
    • Rule-of-5
    • Surface Tension

  • Absolv

    Absolv Module

    The Absolv prediction module calculates Abraham solvation parameters from structure.

    Learn more


  • Blood Brain Barrier Permeation

    Blood Brain Barrier Permeation Module

    The blood-brain barrier (BBB) permeation model in ACD/Labs software provides a comprehensive evaluation of the permeation potential of candidate compounds. While prediction cannot replace experimentation, this module allows compounds to be ranked according to their passive transport across the BBB, based on the following information:

    • Predictions of:
      • Rate of passive diffusion/permeability (logPS)
      • Extent of BBB permeation (logBB)—steady-state distribution ratio of a compound between brain tissue and plasma
      • Brain/plasma equilibration rate (PS * fu, brain)
    • Alerts for compounds likely to undergo transport across the BBB barrier by carrier mediated mechanisms
  • Cytochrome P450 Inhibitors

    Cytochrome P450 Inhibitors Module

    • Calculates the probability that your compound will be an inhibitor of one of the five major drug metabolizing enzymes—CYP3A4, CYP2D6, CYP2C9, CYP2C19, and CYP1A2—at two different IC50 thresholds
      • IC50 < 50 µM (general inhibition);
      • IC50 < 10 µM (efficient inhibition)
    • The model can be trained with experimental data for new compounds in order to expand its applicability domain
  • Cytochrome P450 Substrates

    Cytochrome P450 Substrates Module

    • Calculates the probability that your compound will be a substrate of one of the five major drug metabolizing enzymes—CYP3A4, CYP2D6, CYP2C9, CYP2C19, and CYP1A2
    • The model can be trained with experimental data for new compounds in order to expand its applicability domain
  • Distribution

    Distribution Module

    • Estimates the strength of drug binding to human plasma proteins as either the overall percentage bound in plasma or as an affinity constant to human serum albumin
    • Both the %PPB and logKa(HSA) models are trainable with user data
    • Predict their apparent volume of distribution (Vd)
  • Maximum Recommended Daily Dose

    Maximum Recommended Daily Dose Module

    • Approximate an estimation of the maximum oral dose of drug that can be used in the clinic
  • Oral Bioavailability

    Oral Bioavailability Module

    The oral bioavailability model uses a combination of probabilistic and mechanistic modeling techniques to predict oral bioavailability from structure, and relies on a number of other ACD/Labs prediction algorithms and experimental data sets. Results are provided as a quantitative prediction of bioavailability after oral administration (%F) of a dose defined by the user.

    • Predict a number of endpoints that affect oral bioavailability:
      • Solubility (dose/solubility ratio)
      • Stability in acidic media
      • Intestinal membrane permeability by passive or active transport (with a summary of transporters where relevant)
      • P-gp efflux
      • First pass metabolism in the liver
    • View up to 5 of the most similar structures from the internal training set, with experimental results and literature references
  • Passive Absorption

    Passive Absorption Module

    Human Intestinal Absorption (HIA) and solubility are two key factors that affect oral bioavailability. The Passive Absorption model predicts the human intestinal permeability of drugs, taking into account trans-cellular and para-cellular routes, and ionization-specific differences in permeation rates. Predictions are based on mechanistic models that use a number of physicochemical parameters, including lipophilicity and ionization, as inputs. The model outputs the following calculated parameters:

    • The extent of Human Intestinal Absorption (HIA) in terms of passive transport across the intestine(not affected by any side processes such as limited solubility/dissolution, variable oral dose, chemical stability, active transport, and first pass metabolism in gut or liver), indicating percentage contribution from transcellular and paracellular route.
    • Passive permeability across jejunal epithelium, also indicating absorption rate.
    • Passive permeability across Caco-2 cell monolayers, indicating percentage contribution from transcellular and paracellular route.
  • P-gp Specificity

    P-gp Specificity Module

    P-glycoprotein (P-gp) is a clinically relevant efflux transporter that extrudes compounds from a large variety of cells. Its function has been associated with the drugs' absorption, distribution, excretion, CNS effects, multidrug resistance (MDR). P-gp transports a variety of natural compounds and drugs of different therapeutic areas.

    Rapid identification of drug candidates that are P-gp substrates and/or inhibitors is possible using P-gp specificity model. Filtering and exclusion of P-gp substrates/inhibitors from huge 'in-house' libraries of synthesized compounds or virtual libraries is possible, followed by exclusion of such compounds from further development. P-gp specificity model may serve as an initial screen that could replace screening test based on P-gp ATPase activity measurements and partially replace expensive experiments with P-gp expressing cell monolayers and P-gp knock-out animals.

  • PK Explorer

    PK Explorer Module

    • Estimates a number of parameters determining the pharmacokinetic profile of your compounds by using a set of differential equations from a multi-compartment model describing the organism of an average statistical human:
      • Cp(T)
      • Tmax and Cp(max)
      • AUC after oral and intravenous administrations
      • Oral Bioavailability
  • Regioselectivity of Metabolism

    Regioselectivity of Metabolism Module

    The CYP Regioselectivity model is able to provide valuable insights into a compound's metabolic profile early in the drug discovery process when little or no experimental information is available, and labor intensive investigation of each compound in the screening process is prohibited by the large number of compounds involved.

    • Predict metabolic soft spots for metabolism by:
      • Human liver microsomes (HLM)
      • Five major Cytochrome P450 enzymes (CYP3A4, CYP2D6, CYP2C9, CYP2C19, and CYP1A2)
    • Identify metabolic sites on new chemical entities
    • Guide synthesis of compounds with improved metabolic properties
    • Help identify and elucidate likely metabolite structures


  • Acute Toxicity

    Acute Toxicity Module

    • Predicts quantitative LD50 values for two rodent species after different administration routes:
      • LD50 in mice after oral administration
      • LD50 in mice after intravenous administration
      • LD50 in mice after intraperitoneal administration
      • LD50 in mice after subcutaneous administration
      • LD50 in rats after oral administration
      • LD50 in rats after intraperitoneal administration
    • Estimates qualitative OECD Hazard categories
    • The expert system identifies hazardous substructures potentially responsible for the toxic effect
  • Aquatic Toxicity

    Aquatic Toxicity Module

    • Predict LC50 values of your compounds for two aquatic organisms—fathead minnows (P. promelas) and water fleas (D. magna)
  • Endocrine System Disruption

    Endocrine System Disruption Module

    • Predicts the relative binding affinity of your compounds to the Estrogen Receptor which is associated with the possibility of reproductive toxicity and cancers
  • Mutagenicity

    Mutagenicity Module

    • Provides predictions for the probability of your producing a positive Ames test outcome
  • Health Effects

    Health Effects Module

    • Predicts the probable adverse effects of a compound on particular organs or organ systems based on long term organ specific toxicity studies encompassing various species and routes of administration
    • The following organs and organ systems are considered:
      • Blood
      • Cardiovascular system
      • Gastrointestinal system
      • Kidney
      • Liver
      • Lungs
  • hERG Inhibition

    hERG Inhibition Module

    • Evaluates your compounds for cardiotoxicity related to drug interactions with the human ether-a-go-go (hERG) channel
  • Irritation

    Irritation Module

    • Calculates the potential of a compound to cause moderate or stronger eye and skin irritation as per the Draize test

  • Impurity Profiling Suite

    Impurity Profiling Suite

    Resulting from a collaboration with the FDA, this module offers insights into 21 toxic endpoints reflecting various mechanisms of hazardous activity including:

    • Mutagenicity (Ames test, Mouse Lymphoma Assay, and other standard assays)
    • Clastogenicity (Micronucleus test, Chromosomal Aberrations)
    • DNA damage (Unscheduled DNA Synthesis)
    • Carcinogenicity (FDA rodent carcinogenicity data)
    • Endocrine disruption mechanisms (estrogen receptor binding)

    Learn more

* Available as a discrete bundle
Trainable (machine learning) modules
Purchased as an individual module

Molecular Property Prediction Features

Predict molecular properties from structure, name, or SMILES string.

Easily evaluate results from physicochemical, ADME, and toxicity calculators—each module offers prediction—specific information and tools such as structure highlighting and calculation protocols. Powerful graphing, sorting, and filtering tools further aid evaluation of predicted results.

Use reliability index, probability and/or similar structures from the training set to assess confidence in the predicted result and relevance to your current project.

Apply calculated molecular property data to investigate structural modification/lead optimization to reach the target product profile (absorption, distribution, metabolism, excretion).

Train with experimental data—better reflect proprietary chemical space and improve prediction accuracy using inbuilt machine learning capabilities.

Add custom models and in-house prediction algorithms to core Percepta modules by connecting to an existing web service using an XML protocol, or in the form of a DLL.

  • Evaluate accuracy of prediction with features such as reliability index, and experimental values for similar structures in the database
  • Sub-structure highlighting indicates contributing structural features (Acute Toxicity)
    Sub-structure highlighting indicates contributing structural features
  • Modify lead structures for an optimal property profile (Structure Design)
    Modify lead structures for an optimal property profile

Deployment Options for Molecular Property Calculations on the Percepta Platform

ACD/Labs offers a number of deployment options for our physicochemical, ADME, and toxicity predictors:

Desktop/Thick client

Software installations for individual computers with a graphical user interface. Full physicochemical, ADME and toxicity calculator modules are available (with training capabilities) including the PhysChem Profiler bundle.


Screen tens of thousands of compounds with minimal user intervention—compatible with Microsoft Windows and Linux operating systems (OS). Plug-in to corporate intranets or workflow tools such as Pipeline Pilot.

Percepta Portal/Thin client

Web-based application for prediction of molecular properties (PhysChem, ADME, and toxicity) and data analysis. KNIME integration components available.

Host on your corporate intranet or the cloud. Available for Linux and Windows OS.


This application note discusses the importance of using the machine learning (model training) capabilities of predictive models to improve accuracy.
Read more

Procter & Gamble employs the Percepta platform to replace physicochemical property values derived from physical laboratory experiments, with in silico predictions. Read how Percepta helps chemists and formulators work more efficiently, and assists in decision-making.
Watch Now

Comparison of experimental and published pKa values for 88 cephalosporin antibiotics with ACD/pKa and other prediction software.
Read more

Percepta Portal is a scalable web application that may be easily integrated into your in-house environment. It offers the power of parallel computing, and effortless maintenance and deployment.
Read more

We have found that a number of common misconceptions exist about validated environments. Many arise because previous deployments of software accompanied the installation of new hardware, or have involved informatics systems that are the source of data and reports submitted directly to regulatory authorities. Here we clear up some of the grey areas that seem to have become industry myths.
Read more

Our Virtual Symposium features presentations discussing how ACD/Labs software is used in academia to support research efforts with data processing and property prediction software, and provide hands-on learning and remote teaching with simulation tools.
Watch Recordings

Machine Learning

Thousands of scientists are already benefiting from the machine learning capabilities of ACD/Labs' molecular property calculators and predictors.

Use curated experimental data to expand the training database and the applicability domain of the models. Trainable modules enable the relevance of ACD/Labs’ algorithms to be expanded to novel chemical space not represented in any commercially available products.

Model training and application of training data does not require an understanding of programming and is easily accessible and applicable.

What to look for in molecular property calculators

Broad and/or relevant chemical space coverage—when working with known chemicals choose predictors with broad chemical space coverage. If you are working with novel compounds they will not be represented in datasets of commercially available calculators. Some models may cover more relevant chemical space, however. Evaluate software with compounds for which you have reliable experimental data.

Trainability—if you are working in a niche chemical space or with patent protected compounds look for prediction algorithms that you can train with reliable experimental data—those with machine learning built in.

Capability to assess prediction reliability and accuracy—no calculator will give perfect results for every compound. Choose software that gives you information about the structures/chemical features on which the result is based, or provides an indication of the accuracy of the result (or calculation error).

Why use molecular property calculators and predictors?

The behavior of a molecule in the human body, or any environment, is governed by the molecular properties of that structure. PhysChem and ADME predictors and calculators enable scientists to anticipate how chemical structure affects the performance of a molecule as a drug, pesticide, herbicide, or pollutant etc. Toxicity prediction helps scientists assess potential risk.

In R&D, calculators and predictors of molecular properties help:

  • Reduce the scope and number of experiments
  • Find better novel compounds
  • Shorten the design phase
  • Increase pre-clinical success rates