Chemaxon Descriptor Generation Toolkit

Chemaxon offers predictors covering key ADMET liabilities, so you can filter out weak candidates before synthesis.

• CNS (Central Nervous System) MPO score is a composite metric combining MW, logP, logD, TPSA, HBD, and basic pKa to estimate how likely a compound is to succeed as a CNS drug.
• Blood-brain barrier (BBB) Score is a composite metric based on aromatic ring count, heavy atom count, MWHBN (a descriptor calculated from the molecular weight and the number of H-bond donor and acceptor atoms), TPSA, and pKa that predicts whether a candidate is likely to cross the blood-brain barrier.
• Chemaxon’s hERG predictor uses machine learning models to estimate hERG activity and classify compounds as SAFE or TOXIC, flagging risky scaffolds early in design.

Elemental Analysis provides fundamental molecular properties for general chemistry workflows and analytical scientists.

• Exact and nominal mass, molecular weight and atom count
• Elemental composition
• Molecular formulae
• Molecular composition
• Mass spectrum (isotope distribution) calculation

The partitioning of drugs in microscopic environments (e.g.: lipid bilayers of biological membranes) of different lipophilicity and hydrophilicity heavily influences its pharmacokinetic behavior. The partitioning is described by the logP and logD values of a drug, which are major descriptors in predicting ADMET properties.

The Partitioning calculator provides:

• LogP values using scientifically sound methods
• LogD (pH vs. logD) plot
• HLB (Hydrophilic Lipophilic Balance) number

Solubility in water (commonly referred to as logS) is one of the most important parameters to achieve for desired pharmacological response.

As the majority of drugs are molecules that have ionizable groups, the solubility of the compound depends highly on the pH environment. The Solubility Predictor is able to predict:

• Intrinsic (thermodynamic) solubility in water
• pH-dependent solubility
• A qualitative solubility category

Nuclear Magnetic Resonance spectroscopy is the primary experimental technique to confirm and determine the chemical structure of organic compounds.

As NMR spectrometers are relatively expensive, predicting NMR spectra for a set of possible structures and comparing them with experimental data is a well-established approach to facilitate structure elucidation. The NMR Predictor is able to:

• Predict ¹³C and ¹H NMR spectra for molecules composed of the most frequent elements (H, C, N, O, F, Cl, Br, I, P, S, Si, Se, B, Sn, Ge, Te and As)
• Consider H-H, H-F and C-F couplings in NMR spectrum calculation
• Differentiate between diastereotopic protons

Structural Calculations provide:

• Hydrogen Bond Donor/Acceptor (HBDA) count
• 2D topological descriptors
• 3D geometrical descriptors
• Molecular surface calculations
• 3D conformer generation, molecular dynamics
• 3D Alignment
• Different electronic property calculations

The ability to deal with molecular structures that have identical atomic composition, but different arrangement of bonds or spatial orientation is crucial in the identification, search and property calculation of compounds. Tautomerism, stereoisomerism and resonance are examples of isomerism.

Tautomerization can affect identification, searching and physico-chemical properties of molecules. Stereoisomers have the same physico-chemical properties (e.g.: melting point, solubility), but they differ in pharmacokinetic and pharmacodynamic behavior. Resonance describes the rearrangement of delocalized electrons in the molecule, which results in a set of contributing structures of the original structure.

The Isomer calculator can:

• Generate different tautomer sets useful for searching and tautomer handling
• Calculate tautomer distribution and major tautomer form in water
• Generate tetrahedral and double-bond stereoisomers starting either from a 2D or 3D structure
• Generate resonance structures

Ionization has an essential impact on compound behavior in nearly all protic environments. On top of its application in general and analytical chemistry, relationships to drug pharmacokinetics and pharmacodynamics are well-established. Acidic and basic dissociation constants (pKa) and the microspecies distribution at different pH values are important to quantitatively describe ionization and comprehend physical and biological processes.

The Protonation calculator includes:

• Highly-accurate calculation of pKa values along with pH dependent distribution of relevant microspecies in water
• Calculation of the major microspecies form at a given pH value
• Isoelectric point calculation

Molecular fingerprints are bit vectors encoding of a molecule’s structural features, used to compare, search, and cluster compounds, or to build predictive machine-learning models. Chemaxon offers the following fingerprints:

• Extended Connectivity Fingerprint (ECFP)
• Chemical Hashed Fingerprint (CF)
• Pharmacophore Fingerprint (PF)
• Reaction Fingerprint (RF)

All toolkit components are available through multiple programming interfaces, including Java, Python, REST API (JChem Microservices), .NET, and a command-line interface (cxcalc). Purchased toolkit components are also available for use in Instant JChem, JChem for Office and KNIME with the appropriate additional licenses.