On Two Novel Parameters for Validation of Predictive QSAR Models

Roy, Partha Pratim; Paul, Somnath; Mitra, Indrajit; Roy, Kunal

doi:10.3390/molecules14051660

Cited by 477 publications

(103 citation statements)

References 45 publications

(61 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Y vector randomization 25 times yielded poor Q 2 and R 2 values in our case. R 2 p based on Y-randomization was also applied to the generated model which applied a penalty to the models' coefficient of variance for the dissimilarity between mean coefficient of determination (R 2 r ) of randomized models and coefficient of determination (R 2 ) of the non-randomized model [31]. In order to have a superior predictive potential of the model, a modified R 2 (R 2 m ) was determined [31].…”

Section: Methodsmentioning

confidence: 99%

“…R 2 p based on Y-randomization was also applied to the generated model which applied a penalty to the models' coefficient of variance for the dissimilarity between mean coefficient of determination (R 2 r ) of randomized models and coefficient of determination (R 2 ) of the non-randomized model [31]. In order to have a superior predictive potential of the model, a modified R 2 (R 2 m ) was determined [31]. Further, the model was externally validated by predicting the activity of external set or test set of compounds and evaluated by means of coefficient of determination (R 2 pred ), standard error of external prediction (SEP), and the average relative error (ARE pred ) [32].…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Benzo[e]pyrimido[5,4-b][1,4]diazepin-6(11H)-one derivatives as Aurora A kinase inhibitors: LQTA-QSAR analysis and detailed systematic validation of the developed model

et al. 2015

View full text Add to dashboard Cite

Aurora kinases are sub-divided into Aurora A, Aurora B, and Aurora C kinases that are considered as prospective targets for a new class of anticancer drugs. In this work, a 4-D-QSAR model using an LQTA-QSAR approach with previously reported 31 derivatives of benzo[e]pyrimido[5,4 -b][1,4]diazepin -6(11H)-one as potent Aurora kinase A inhibitors has been created. Instead of single conformation, the conformational ensemble profile generated for each ligand by using trajectories and topology information retrieved from molecular dynamics simulations from GROMACS package were aligned and used for the calculation of intermolecular interaction energies at each grid point. The descriptors generated on the basis of these Coulomb and Lennard-Jones potentials as independent variables were used to perform a PLS analysis using biological activity as dependent variable. A good predictive model was generated with nine field descriptors and five latent variables. The model showed [Formula: see text]; [Formula: see text] and [Formula: see text]. This model was further validated systematically by using different validation parameters. This 4D-QSAR model gave valuable information to recognize features essential to adapt and develop novel potential Aurora kinase inhibitors.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Benzo[e]pyrimido[5,4-b][1,4]diazepin-6(11H)-one derivatives as Aurora A kinase inhibitors: LQTA-QSAR analysis and detailed systematic validation of the developed model

et al. 2015

View full text Add to dashboard Cite

show abstract

“…Octanol/water partition coefficients and related maximum residue limits (MRL) in fat tissue LogP  arithmetic mean of octanol/water partition coefficient; LogD7.4  arithmetic mean of octanol/buffer pH 7.4 partition coefficient; CLogD7.4 -octanol/buffer pH 7.4 partition coefficient (MarvinSketch algorithm); CLogP -octanol/water partition coefficient (MarvinSketch algorithm); XLogP -octanol/water partition coefficient (ALOGPS algorithm, PubMed); LogKow -octanol/water partition coefficient (EPI Suite TM algorithm) (16) ) of the observed versus predicted data was obtained. The leave-one-out (LOO) method was used for model cross-validation (14,27). Squared cross-validated correlation coefficient (Q 2 ) parameter and differences between Q 2 and R 2 were calculated as a measure of the internal performance and model predictive ability.…”

Section: Methodsmentioning

confidence: 99%

Indirect relationship between lipophilicity and maximum residue limit of drugs determined for fatty tissue

Marczak¹,

Okoniewska²,

Okoniewski³

et al. 2015

Bulletin of the Veterinary Institute in Pulawy

View full text Add to dashboard Cite

The aim of this study was to determine the correlation between lipophilicity and maximum residue limit (MRL) value specified for veterinary drugs in the fatty tissue of various animal species. The analysis was performed on a group of 73 compounds with different modes of action and MRL values determined for the fatty tissue of animals. Additionally, the logarithm of water/organic phase partition ratio (LogP) and the ratio of ionised and unionised substance in buffer with pH 7.4 (LogD7.4) were calculated. The main analysis was performed after the division of the whole group into six fractions. The linear correlation and regression analysis were used to determine the indirect relationship between the mean arithmetic value of LogP or LogD7.4 in selected fractions and related LogMRL of the drugs tested. The calculations revealed a linear correlation between fractioned lipophilicity and LogMRL values for the analysed compounds. The existence of indirect relationship between lipophilicity and MRL values determined for fatty tissue was confirmed.

show abstract

“…A parameter R 2 p was introduced to evaluate the Y-randomization results. The parameter penalized the model for the difference between squared mean correlation coefficient (R 2 r ) of randomized models and squared correlation coefficient (R 2 ) of the non-randomized model [55] and was calculated by the following equation:…”

Section: Comsia Models Generationmentioning

confidence: 99%

“…To further validate the predictive ability of the model, a modified r 2 (r 2 m (overall) ) was introduced as the following equation [55]:…”

Section: Comsia Models Generationmentioning

confidence: 99%

3D-QSAR modeling and molecular docking study on Mer kinase inhibitors of pyridine-substituted pyrimidines

Zhuang

et al. 2014

Mol Divers

View full text Add to dashboard Cite

Mer kinase is a novel therapeutic target for many cancers, and overexpression of Mer receptor tyrosine kinase has been observed in several kinds of tumors. To deeply understand the structure-activity correlation of a series of pyridine/pyrimidine analogs as potent Mer inhibitors, a combined molecular docking and three-dimensional quantitative structure-activity relationship modeling was carried out. A comparative molecular similarity indices analysis model was developed based on the maximum common substructure alignment. The optimum model exhibited statistically significant results: the cross-validated correlation coefficient q2 was 0.599, and non-cross-validated r2 value was 0.984. Furthermore, the results of internal validation such as bootstrapping, Y-randomization as well as external validation (the external predictive correlation coefficient r2 ext = 0.728) confirmed the rationality and good predictive ability of the model. Using the crystal structure of Mer kinase, the selected pyridine/pyrimidine compounds were docked into the enzyme active site. Some key amino acid residues were determined, and hydrogen bonding and hydrophobic interactions between Mer kinase and inhibitors were identified. The satisfactory results from this study may aid in the research and development of novel potent Mer kinase inhibitors.

show abstract

On Two Novel Parameters for Validation of Predictive QSAR Models

Cited by 477 publications

References 45 publications

Benzo[e]pyrimido[5,4-b][1,4]diazepin-6(11H)-one derivatives as Aurora A kinase inhibitors: LQTA-QSAR analysis and detailed systematic validation of the developed model

Benzo[e]pyrimido[5,4-b][1,4]diazepin-6(11H)-one derivatives as Aurora A kinase inhibitors: LQTA-QSAR analysis and detailed systematic validation of the developed model

Indirect relationship between lipophilicity and maximum residue limit of drugs determined for fatty tissue

3D-QSAR modeling and molecular docking study on Mer kinase inhibitors of pyridine-substituted pyrimidines

Contact Info

Product

Resources

About