Multiple Pharmacokinetic Parameter Prediction for a Series of Cephalosporins

Turner, Joseph V.; Maddalena, D.J.; Cutler, David J.; Agatonović-Kuštrin, Snežana

doi:10.1002/jps.10314

Cited by 29 publications

(20 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sensitivity is defined as the ratio of error of a retrained optimum model, which does not contain the information of a specific molecular descriptor, to the error of the optimum model that includes the information from the molecular descriptor [25]. Since ANNs compute the output as a sum of nonlinear transformations of linear combinations of the inputs, sensitivity shows the percentage contribution of a corresponding input to the output value and reveals the effect that a change in that particular input has on output.…”

Section: Selection Of Input Variablesmentioning

confidence: 99%

The Use of Fourier Transform Infrared (FTIR) Spectroscopy and Artificial Neural Networks (ANNs) to Assess Wine Quality

Agatonović-Kuštrin¹,

Morton²,

Yusof³

2013

Mod Chem Appl

View full text Add to dashboard Cite

The aim of this study was to develop a simple method to assess wine quality from its Fourier Transform Infrared Spectroscopy (FTIR) spectrum with minimal or no sample preparation. FTIR spectral data of selected wine samples, grape variety, wine barrel type, wine type and production year were correlated with total phenolic content, total and volatile acidity and alcohol content using Artificial Neural Networks (ANNs). A total of 20 (2 whites and 18 reds) different wines used in this study came from three different states across Australia; New South Wales, Victoria and South Australia.FTIR spectroscopy proved to be a promising technique that provides a rapid and accurate method in the quality assessment of wine. A plot of the values predicted by the validated ANN models showed excellent correlation with the experimentally measured values for acetic acid concentration, alcohol content, total phenols, and total acidity (r=0.898-0.942).

show abstract

Section: Selection Of Input Variablesmentioning

confidence: 99%

The Use of Fourier Transform Infrared (FTIR) Spectroscopy and Artificial Neural Networks (ANNs) to Assess Wine Quality

Agatonović-Kuštrin¹,

Morton²,

Yusof³

2013

Mod Chem Appl

View full text Add to dashboard Cite

show abstract

“…The models are derived by different statistical and machine learning methods as artificial neural networks (ANN) (7,9,13), multiple linear regression (MLR) (8,10,12,14,15,18,19), partial least squares (PLS) (10 -12, 16, 18, 20), Bayesian neural networks (BNN) (16), classification and regression trees (CART) (16), mixed determinant analysis -random forest (MDA -RF) (17), recursive partitioning classification (RPC) (20).…”

Section: Introductionmentioning

confidence: 99%

Quantitative Structure – Pharmacokinetics Relationships Analysis of Basic Drugs: Volume of Distribution

2015

View full text Add to dashboard Cite

-Purpose. The early prediction of pharmacokinetic behavior is of paramount importance for saving time and resources and for increasing the success of new drug candidates. The steady-state volume of distribution (VD ss ) is one of the key pharmacokinetic parameters required for the design of a suitable dosage regimen. The aim of the study is to propose a quantitative structure -pharmacokinetics relationships (QSPkR) for VD ss of basic drugs. Methods: The data set consists of 216 basic drugs, divided to a modeling (n = 180) and external validation set (n = 36). 179 structural and physicochemical descriptors are calculated using validated commercial software. Genetic algorithm, stepwise regression and multiple linear regression are applied for variable selection and model development. The models are validated by internal and external test sets. Results: A number of significant QSPkRs are developed. The most frequently emerged descriptors are used to derive the final consensus model for VD ss with good explanatory (r 2 0.663) and predictive ability (q 2 LOO-CV 0.606 and r 2 pred 0.593). The model reveals clear structural features determining VD ss of basic drugs which are summarized in a short list of criteria for rapid discrimination between drugs with a large and small VD ss . Conclusions: Descriptors like lipophilicity, fraction ionized as a base at pH 7.4, number of cycles and fused aromatic rings, presence of Cl and F atoms contribute positively to VD ss , while polarity and presence of strong electrophiles have a negative effect.

show abstract

“…A wide suite of statistical and machine learning methods were used: multiple linear and non-linear regression analysis (3), partial least squares (4 -12), artificial neural networks (13), k-nearest neighbor (9,10,14), general regression network and support vector regression (10). The models show satisfactory predictive ability, but some of them are not well validated, and another do not offer sufficient interpretability.…”

Section: Introductionmentioning

confidence: 99%

Quantitative Structure – Pharmacokinetic Relationships for Plasma Clearance of Basic Drugs with Consideration of the Major Elimination Pathway

Zhivkova

2017

J Pharm Pharm Sci

View full text Add to dashboard Cite

Purpose. The success of a new drug candidate is determined not only by its efficacy and safety, but also by proper pharmacokinetic behavior. The early prediction of pharmacokinetic parameters could save time and resources and accelerate drug development process. Plasma clearance (CL) is one of the key determinants of drug dosing regimen. The aim of the study is development of quantitative structure -pharmacokinetics relationships (QSPkRs) for the CL. Methods. A dataset consisted of 263 basic drugs, which chemical structures were described by 154 descriptors. Genetic algorithm, stepwise regression and multiple linear regression were used for variable selection and model development. Predictive ability of the models was assessed by internal and external validation. Results. A number of significant QSPkR models for the CL were derived with respect to the primary elimination pathway (renal excretion, metabolism, or CYP3A4 mediated biotransformation), as well for the unbound clearance (CL u ). The models were able to predict 52 -80% of the drugs from external validation sets within the 2-fold error of the experimental values with geometric mean fold error 1.57 -2.00. Conclusions. Plasma protein binding was the major restrictive factor for the CL of drugs, primarily cleared by metabolism. The clearance was favored by lipophilicity and several structural features like OH-groups, aromatic rings, large hydrophobic centers, aliphatic groups, connected with electro-negative atoms, and non-substituted aromatic Catoms. The presence of Cl-atoms and abundance of 6-member aromatic rings or fused rings had negative effect. The presence of ether O-atoms contributed negatively to the CL of both metabolism and renally excreted drugs, and urine excretion was favored by the presence of 3-valence N-atoms. These findings give insight on the main structural features governing plasma CL of basic drugs and could serve as a guide for lead optimization in the drug development process.

show abstract

Multiple Pharmacokinetic Parameter Prediction for a Series of Cephalosporins

Cited by 29 publications

References 38 publications

The Use of Fourier Transform Infrared (FTIR) Spectroscopy and Artificial Neural Networks (ANNs) to Assess Wine Quality

The Use of Fourier Transform Infrared (FTIR) Spectroscopy and Artificial Neural Networks (ANNs) to Assess Wine Quality

Quantitative Structure – Pharmacokinetics Relationships Analysis of Basic Drugs: Volume of Distribution

Quantitative Structure – Pharmacokinetic Relationships for Plasma Clearance of Basic Drugs with Consideration of the Major Elimination Pathway

Contact Info

Product

Resources

About