[ properties | Number of Atoms | Tox Score | Blood Brain Barrier Score | MolSynth Score | LogD | PSA ]
| Video |
To calculate chemical properties for compounds within a chemical table:
- Read in the chemical table.
- Select Chemistry/Calculate Properties and a window as shown below will be displayed.
- Arguments - usually the column with the chemical structure is called mol. If this is not the case click on the drop down button and select the correct column.
- New column location - select where you would like to insert the new columns.
- New column name - This option is usually greyed out but the column name will be the same as the property name.
- Select the properties you wish to calculate using the 'tick' check boxes.
- Click OK and the properties will be added as new columns in the chemical table.
Actions Button Helps with Selections
If you click on the drop down button labeled Actions you will see 3 selection options:
- Select All properties.
- Uncheck selected options.
- Remove unchecked.
A wide range of ADME properties can be quickly and accurately calculated on large chemical datasets. This tool enables you to flag chemicals that might have poor ADME-TOX properties before experimental testing.
Some of the properties you can predict are:
- MolWeight - Molecular weight from .mol
- MolFormulaChemical - formula,e.g. C2H6O, from .mol
- IupacName - IUPAC nomenclature name from .mol
- MolLogP - Octanol water partition, -Log(C_w/C_oct) from .mol
- MolLogS- Water solubility -Log(C_aggr) from .mol 0 and higher: Highly soluble 0 to −2: Soluble −2 to −4: Slightly soluble Less than <−6: Insoluble
- MolPSA - Polar surface area from .mol
- MolPSA 3D = Polar surface area after conversion to 3D
- MolArea - Total surface area from .mol
- MolVolume- Molecular volume from .mol
- MoldHf - Heats of formation from elements from .mol
- MolhERG - hERG binding prediction
- MolHalfLife - Half life, hours
- MolPAINS" Pan Assay Interference Compounds
- MolCACO2 - Predict CACO-2 Permeability LogP
- MolLD50 - Predict LD50 in mg/kg
- MolPAMPA - Predict PAMPA Permeability
- MolPGPINHIB - Predict P=Glycoportein Inhibition
- MolPGPSUBST - Predict P-Glycoprotein substrate
- DrugLikeness Empirical drug-likeness [-1:+1] from .mol
- Tox Score - Chemical Alert Collected from Chemical supplier and other sources
- Smiles SMILES/SMARTS: string notation of chemical or chemical patterns derived from .mol
- Atom Counts - Atom counts
- Bond Counts - Bond counts
- Topological Descriptors - Toplogical, connectivity and shape indices
- BadGroups - Unwanted or reactive chemical functionality from .mol
- Covalent/Prodrug Groups - Potential chemical groups that can be linked covalently or cleaved in prodrug.
- Nof_Atoms - Number of atoms from .mol
- Nof_Molecules - Number of individual molecules in .mol drawings
- Nof_Fragments - Number of fragments
- Nof_Chirals - Number of chiral centers, R,S,or (RS) from .mol
- Nof_RingsNumber of rings in the SSSR from .mol
- Max_Ring_Size - Largest independent ring size from .mol
- Min_Ring_Size - Smallest independent ring size from .mol
- Max_Fused_Rings - Number of elementary rings in the largest fused ring from .mol
- Nof_RotB - Number of freely rotatable bonds from .mol
- Nof_HBA - Number of hydrogen bond acceptors from .mol
- Nof_HBD - Number of hydrogen bond donors from .mol"
- MolCharge - Calculate total formal charge at given Ph
- pKa of the Most Basic Group - Calculate pKa of most basic group
- pKa of the Most Acidic Group - Calculate pKa of most acidic group
- InChI - Calculate InChi string
- InChIKey - Calclate InChI key string
- PubChem CID = RetrievePubChem CID by structure
- Name or ID from Structure - Convert Structure to Name/CAS o PubChem CID
- ChEMBL Bioactivity - ChEMBL Bioacticity from Structure
- Drug Bank ID - Retrieve Drug Bank ID number
- Structure to Name - Tanslate chemical structure to Name, CAS, DrugBank, ID, ChEMBL ID
- FindChemPatterns - Annotate .mol drawings by found chemical substructure
- 2Dfrom_Smiles - Convert SMILES to 2D .mol chemical drawings
- 2Dfrom_InChi - Convert InChi to 2D .mol chemical drawings
You can calculate the Number of Atoms using:
- Chemistry/Calculate Properties
- Choose the Function Chemical from the drop down list and then Nof_Atoms.
- Add the SMARTS string for the atoms you are trying to count (e.g. sp3 : [*;^3] halogen : X or [F,Cl,Br,I])
Nof SP3 example:
The ToxScore gives a prediction for how reactive or toxic a chemical is.
How the ToxScore is Calculated
Around 1000 SMARTS strings associated with toxicity/reactivity were collected from various sources. They were assigned demerit score based on their perceived toxicity and frequency of appearance in approved drugs. The toxscore of any compound was calculated by summing up all the matching SMARTS strings. A toxscore >= 1. indicates likely toxicity based on substructure match.
How to Calculate ToxScore
- Read into ICM an SDF file. A chemical spreadsheet will be displayed in the GUI.
- insert a column into the chemical spreadsheet. Right click on a column header > choose insert column > choose the chemical function > select Tox Score.
- A column containing a Tox Score and Tox Names will be displayed.
10.11.4 Blood Brain Barrier Score |
The Blood Brain Barrier prediction score has been developed based on the method described By Gupta et al in the Journal of Medicinal Chemistry 2019 - see DOI: 10.1021/acs.jmedchem.9b01220
- Read into ICM an SDF file. A chemical spreadsheet will be displayed in the GUI.
- Chemistry/Calculate Properties and choose the 'bbb' blood brain barrier
- A score will be displayed in a column in the SDF file. A score >4 indicates the chemical can pass the BBB
10.11.5 Synthesizability Score |
Our measure of synthetic accessibility is based on the statistics of frequencies of ECFP4 fingerprints calculated over a large database of synthetic compounds. Fragments/fingerprints with frequencies <10000 (occurring in less than 0.05% of the database compounds) contribute to perceived ‘difficulty' (as in 4-Log(freq)). Functional form is such that accessibility drops as the number and/or rarity of unusual fingerprints in the molecule increases.
A reasonable cutoff between accessible and difficult is 0.5.
LogD models is based on LogP and pKa most basic and most acidic groups (pKa_ma and pKa_mb)
LogD = LogP - Log( 1 + 10^(-7.4 + pKa_mb )) # for pKa_ma > 7.4 and pKa_mb exists
LogD = LogP - Log( 1 + 10^( 7.4 - pKa_ma )) # for pKa_mb < 7.4 and pKa_ma exists
LogD = LogP # for other cases
The PSA model is trained on diverse set of drug-like molecules with 3D conformations from Enamine HTS collection. (~200K of 3D conformations). Training values for PSA is averaged across multiple conformations per compound (calculated using ' show surface mute area' ICM command).
- Descriptors: ECFP4
- Model: PLS
- Internal Cross Validation (leave 1/5 out)
- Q2= 0.98, R2=0.99
|