Dependence and Independence: Structures, Testing, and Measuring
Albert Vexler, Alan D. Hutson, Xiwei Chen in Statistical Testing Strategies in the Health Sciences, 2017
Bolboaca and Jäntschi (2006) studied a sample of 67 pyrimidine derivatives with inhibitory activity on Escherichia coli dihydrofolate reductase (DHFR) by the use of molecular descriptor families on structure–activity relationships. The use of Pearson, Spearman, Kendall, and gamma correlation coefficients in the analysis of structure–activity relationships of biologic active compounds was studied and presented.
Quantitative structure–activity relationship models for compounds with anticonvulsant activity
Published in Expert Opinion on Drug Discovery, 2019
Carolina L. Bellera, Alan Talevi
Quantitative Structure–Activity Relationships (QSAR) have become a cornerstone in the drug discovery and Medicinal Chemistry fields, where they can be used to explain the differences in the activity of congeneric series, identify novel bioactive scaffolds (through virtual screening) and drive the design and optimization of new drug candidates [8–10]. QSAR methods attempt to establish a correlation between molecular features and/or physicochemical properties (numerically captured into molecular descriptors), and biological properties, reflected either as a continuous or a categorical variable [8,9]. Probably, the most popular definition of a molecular descriptor belongs to Consonni and Todeschini, ‘the molecular descriptor is the final result of a logic and mathematical procedure which transforms chemical information encoded within a symbolic representation of a molecule into a useful number or the result of some standardized experiment’ [11].
Lead optimization of 4-(thio)-chromenone 6-O-sulfamate analogs using QSAR, molecular docking and DFT – a combined approach as steroidal sulfatase inhibitors
Published in Journal of Receptors and Signal Transduction, 2021
The energy minimized compounds were subjected to molecular descriptor calculation for calculating 2D descriptors using PaDEL freeware version 2.21 [25]. A large number of descriptors were calculated and the variables were pre-filtered by deleting all the missing values and excluding the zero values to avoid correlation among the descriptors. Furthermore, pairwise correlation was used to filter out the descriptors with more than 0.70 values. From the correlation matrix (Table 2) obtained after filtering all the excluded descriptors, the electrotopological 2D descripotrs BCUTp-1h, SdssC and maxHBa [26] showed greater correlation with activity. The major contributing features toward STS inhibition in a thiochromenone nucleus are hydrophobic and electronic features. The detailed descriptor analysis that contributes for maximum potency is BCUTp-1h which is the nlow highest polarizability weighted BCUTS, SdssC is the sum of atom type E-states for (strong) hydrogen bond acceptors.The best models developed from hydrophobic and E-state descriptors after conducting many trial models are presented here.
Computational modeling of human oral bioavailability: what will be next?
Published in Expert Opinion on Drug Discovery, 2018
Miguel Ángel Cabrera-Pérez, Hai Pham-The
The molecular descriptor is the final result of transforming the chemical information of the molecule into a useful number, by means of mathematical or logical procedures [33]. Different classes of molecular descriptors have been employed to correlate chemical structure and ADME properties [34]. Although thousands of molecular descriptors are available, their selection must be based on basic knowledge of the property to be modeled, in order to develop successful prediction models [10]. In the prediction of oral bioavailability, descriptors with 1D, 2D, and 3D structural information have proven to be useful in deriving robust and predictive models [26].
Related Knowledge Centers
- Partition Coefficient
- Molar Refractivity
- Mathematical Chemistry
- Topological Index
- Quantitative Structure–Activity Relationship
- Applicability Domain
- Chemical Database
- Docking
- Cahn–Ingold–Prelog Priority Rules