Structure-activity relationship ( commonly known as SAR) is an approach to calculate and evaluate the relationship between the molecular or chemical structure and its biological activity. The data that is collected by performing the SAR methodology is stored in a concentrated database known as the Structure-activity relationship database.
Quantitative structure-activity relationship
QSAR models are just like the SAR model. But, it is just a theoretical approach to calculate a quantitative measure of the relationship between a chemical structure and its physical or biological activity. QSAR’s data is not stored inside the Structure-activity relationship database.
Tools for QSAR
Many QSAR tools are well developed that can predict the toxicity level of a chemical substance just by feeding its chemical structure.
Expert system ( commonly known as Rule-based SARs)
This tool is specialized in providing the right prediction and gives structural alerts promptly whenever the toxicity level of a chemical stems high.
It is mostly applied only to the toxic endpoints where the exact mechanism is already known. It is also used on less chemical informative data.
The free expert system tool is Toxtree. If you need more specific data or if you need the expert system tool for commercial use then go for Derek Nexus.
The underlying algorithm that is used in a statistical model is data mining, machine learning, and mathematical models.
The application of these types of tools is commonly on the toxic endpoints with very little or zero knowledge of the mechanism of action. It is also used on a significant amount of data.
Freely available statistical model tools are VEGA, LAZAR, and EPA T.E.S.T. Buy MultiCASE for any commercial purpose.
Rule-based and statistical modeling are the two different types of algorithms that are used in a hybrid model. A hybrid modeled tool is a combination of both the rule-based and the statistical methodologies. This can be used to get high accuracy in the data.
Hybrid modeled tools are only for commercial use. These are TIMES and Catalogic tools.
Needs and Application for QSAR
- QSAR is needed to test ample chemicals to evaluate the toxicity point level. More sensitive chemicals are tested using QSAR methods and tools only. If the vivo data is insufficient then QSAR data is the one that comes to the rescue.
- QSAR is an effective alternative to animal testing. It is also faster and cheaper in application and results when compared to animal testing.
- QSAR data is the one that is used in the absence of experimental data. Experimental animal testing methods could be replaced by QSAR methods.
- These methods are used for effective designing and developing new drugs, colognes, and other chemical substances.
- QSAR methods are less hazardous in comparison to the other methods. Using this method to evaluate data will ensure less chemical risk.
The QSAR Workflow
- The molecular descriptors are first generated from the chemical structure
- After extracting all the molecular descriptors, the most relevant descriptor is chosen.
- These descriptors are mapped to a toxic endpoint
- The model is validated at the fourth stage
- The model is applied
Molecular descriptors in the QSAR workflow
Molecular descriptors are nothing but the quantification of different types of molecular properties in a chemical compound. There are various levels of chemical representation of these descriptors ranging from 1D to 4D.
1D – The properties that are inferred only by the chemical formula of the chemical is known as the 1D descriptor
2D – A 2D structure formula is inferred and the properties of the structure of the chemical are considered.
3D -The spatial shape of the chemical is considered for its one conformation
4D – This is the same as that of the 3D instead it is considered for multiple conformations.
2D descriptors in QSAR workflow
1. Constitutional descriptors
They withhold the properties relevant to the molecular structure of the compound. Molecular weight, number of atomic rings are some of the examples of a constitutional descriptor.
The electrostatic descriptor contains the properties that resemble the electronic nature of the chemical compound. Examples might include partial charges and atomic net
3. Topological descriptors
These kinds of descriptors are obtained by treating the structure of the chemical compound as a graph. An example of a topological descriptor is the total number of bonds in a non-hydrogen atom.
4. Geometrical descriptors
Geometrical descriptors constitute the properties that are related to the spatial arrangements of the compound. Vander Waals Area is a great example of a geometrical descriptor.
5. Fragment-based descriptor
The properties of substructural motifs represent a fragment-based descriptor. Examples might include an MDL key and molecular fingerprints.