Parameter Questions
Last updated
Last updated
These descriptors are calculated using (version 18.1).
The SMILES are calculated using the .
Standard InChI is calculated using the algorithm from IUPAC. The version of InChI used in CMNPD is 1.05.
InChI Key is based on a strong hash (SHA-256 algorithm) of an InChI string.
All physicochemical properties are calculated using algorithms available in .
The sum of the atomic masses with the isotope average used for each atomic mass.
The sum of the atomic masses with the most common isotope is used for each atomic mass.
Number of rotatable bonds in the molecule. Based on matching this SMARTS pattern:
Number of hydrogen bond acceptors: number of heteroatoms (Oxygen, Nitrogen, Sulfur, or Phosphorus) with one or more lone pairs, excluding atoms with positive formal charges, amide and pyrrole-type Nitrogen, and aromatic Oxygen and Sulfur atoms in heterocyclic rings. Based on matching these SMARTS patterns:
Number of hydrogen bond donors: number of heteroatoms (Oxygen, Nitrogen, Sulfur, or Phosphorus) with one or more attached Hydrogen atoms. Based on matching these SMARTS patterns:
The number of aromatic rings in the molecule.
The number of non - hydrogen atoms in the molecule.
The values range from 0 -1 where 1 is the most drug-like and 0 the least drug-like.
Blood Brain Barrier Penetration: Base 10 logarithm of (brain concentration)/(blood concentration)
Blood Brain Barrier Penetration Level: Predicts the blood-brain permeation level based on the following categories
Level
Value
Description
0
Very High
Brain-blood ratio greater than 5:1
1
High
Brain-blood ratio between 1:1 and 5:1
2
Medium
Brain-blood ratio between 1:1 and 5:1
3
Low
Brain-blood ratio less than 0.3:1
4
Undefined
Outside 99 percent confidence ellipse
Intestinal absorption is defined as a percentage absorbed rather than as a ratio of concentrations (cf.blood-brain penetration); a well-absorbed compound is one that is absorbed at least 90 percent into the bloodstream in humans. The levels are defined as:
0 = Good
1 = Moderate
2 = Low
3 = Very low
Aqueous Solubility: The base 10 logarithm of the molar solubility as predicted by the regression.
Aqueous Solubility Level: Assigns the molecule to one of seven solubility classes based on the value of ADMET Solubility. The classes define the solubility relative to a compendium of known drugs:
Level
Value
Description
0
< -8.0
Extremely low solubility, lower than 95 percent of drugs
1
(-8.0, -6.0)
Very low solubility, at border line of 95 percent of drugs
2
(-6.0, -4.0)
Low solubility, at lower end of 95 percent of drugs
3
(-4.0, -2.0)
Good, slight soluble to soluble
4
(-2.0, 0.0)
Optimal solubility
5
> 0.0
Very soluble, perhaps too soluble
Calculated value for the lipophilicity of a molecule expressed as log (octanol/water partition coefficient). Method used for the calculation is as described in "Prediction of physicochemical parameters by atomic contributions" by .
The Molecular Polar Surface Area descriptors are calculated using a method based on the published method: "Fast calculation of molecular polar surface area as a sum of fragment based contributions and its application to the prediction of drug transport properties" by .
This is the quantitative estimate of drug-likeness as described in "Quantifying the chemical beauty of drugs" by .
All ADMET parameters are calculated using ADMET models available in (version 18.1).
The model predicts blood-brain barrier penetration (BBB) after oral administration. Method used for the calculation is as described in "Prediction of drug absorption using multivariate statistics" by . The calculable properties of the component are:
This model predicts human intestinal absorption (HIA) after oral administration. Method used for the calculation is as described in "Prediction of drug absorption using multivariate statistics" by and "Prediction of intestinal permeability" by .
The Aqueous Solubility descriptor uses linear regression to predict the aqueous solubility of each compound in water at 25 degrees Celsius. Method used for the calculation is as described in "Prediction of aqueous solubility of a diverse set of compounds using quantitative structure−property relationships" by .
This model predicts whether a particular compound is an inhibitor of the CYP2D6 isozyme of cytochrome P-450 with a value of True or False This model trained on a data set of compounds as described in "Use of robust classification techniques for the prediction of human cytochrome P450 2D6 inhibition" by .
This model predicts dose-dependent human hepatotoxicity with a value of True (toxic) or False (nontoxic). This model trained on a data set of 436 compounds as described in "In silico models for the prediction of dose-dependent human hepatotoxicity" by .
This model predicts Plasma-Protein binding - whether compound is likely to be highly bound to carrier proteins in the blood. If True, the compound is estimated to be a binder (>=90%). Otherwise, it is estimated to be weaker or non-binder (<90%). This model trained on several data sets of compounds using modified Bayesian learning as described in "Classification of kinase inhibitors using a Bayesian model" .