Untitled

Andrews, C. Webster; Bennett, Lee; Yu, Lin

doi:10.1023/a:1007556711109

Cited by 130 publications

(30 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The adaptive search performed by TOPAS was re-started several times (population size 50 molecules; 100 generations), where the two recurring motifs 2 and 3 were proposed, possibly representing marked optima in search space. We recognized the issue concerning bioavailability based on the high lipophilicity of the proposed designs (clogP 6~7) [17], however, due to the clogP values of other reported CB-1 ligands (seed 1, clogP 5.5; D 1 -THC, clogP 7.6; anandamide, clogP 6.3) we accepted the highly tractable proposed molecular structures 2 and 3 as a starting point for a hit generation initiative.…”

Section: Full Papermentioning

confidence: 99%

Identification of novel cannabinoid receptor ligands via evolutionary de novo design and rapid parallel synthesis

et al. 2004

View full text Add to dashboard Cite

A rapid entry to novel and patentable cannabinoid receptor (CB-1) ligands has been identified using computer-based de novo design in combination with parallel synthesis. Small targeted compound arrays were readily prepared from the designs yielding active hit rates of 6% and 10%, respectively. This represents a hit rate enrichment of up to two orders of magnitude compared with corporate compound collections and is the first report of the application of the TOPAS algorithm for the de novo design of GPCR ligands.G-protein coupled receptors (GPCRs) are by far the largest single group of targets within pharmaceutical research today [1]. Not only are they involved in various pharmacological pathways but are also implicated in many different therapeutic indications ranging from infectious and metabolic diseases to CNS disorders. Within the GPCR protein family, extremely limited biostructural information is available due to inherent difficulties arising from large protein size and high lipophilicity. This makes isolation and biostructural characterization of GPCRs a major challenge currently. Rational drug design based upon structural homology is challenging due to the large size and complexity of the target proteins. This requires both computationally intensive and lengthy molecular dynamics refinement as well as detailed mutagenesis efforts to produce an accurate, validated and predictive model due to our limited understanding of ligandreceptor interaction patterns in GPCRs [2]. Successful examples of this approach have been reported ever toward designing novel cannabinoid receptor ligands by structurebased methods which rely on receptor binding-site models [3]. However in general, ligand discovery efforts are commonly initiated based on a high throughput random screening (HTS) program placing pressure on assay development, compound library management and the corresponding logistics, as key bottlenecks. Since the majority of −drugable× targets within pharmacological research belong to this class of proteins it is clear that there is a need for novel technologies which complement classical HTS approaches [4]. Herein we describe the application of an iterative ligand-based library design strategy for the production of small focused compound collections with the goal of generating novel and patentable ligands for the GPCR cannabinoid subtype receptor 1 (CB-1).With the advent of combinatorial chemistry and ultra-HTS, the generation of large compound libraries has become commonplace within the drug discovery industry. Initially, library designs were simply based on chemical feasibility rather than target relevant criteria. Combinatorial chemistry libraries consisted of unpurified mixtures of compounds submitted directly for biological testing, however due to the high false positive and low validated hit rates, the trend is now increasingly towards purified discretes. More recently, a shift from random compound libraries to sophisticated maximally diverse compound collections and targeted compound arrays was favoured [5]. W...

show abstract

Section: Full Papermentioning

confidence: 99%

Identification of novel cannabinoid receptor ligands via evolutionary de novo design and rapid parallel synthesis

et al. 2004

View full text Add to dashboard Cite

show abstract

“…Up to now, many prediction models of ADMET properties have been established [3,4], but some of these models are still not competent for the prediction purpose [5], such as, typically, oral bioavailability (BIO) [6][7][8] and human plasma protein binding rate (PPBR) [8].…”

Section: Introductionmentioning

confidence: 99%

Prediction models of human plasma protein binding rate and oral bioavailability derived by using GA–CG–SVM method

Yang

Zhang

et al. 2008

Journal of Pharmaceutical and Biomedical Analysis

View full text Add to dashboard Cite

“…In order to comparison these models, standard errors (SE) and coefficient of multiple determination (R) calculated from below equation. (2) and (3) where y i and were the experimental, predicted and average experimental values of oral bioavailability, respectively and n is the number of data. The statistical parameters of these models are shown in Table 3.…”

Section: Modelingmentioning

confidence: 99%

“…For example, Andrew et al proposed a QSAR model that can achieve the coefficient of multiple correlation of R 2 = 0.71 by using 85 molecular descriptors. 2 In 2004, Tuner et al developed QSAR model for prediction of oral bioavailabilities of some chemicals by using the artificial neural network (ANN). 4 The training and testing correlation coefficients given by the model were 0.736 and 0.897, respectively.…”

Section: Introductionmentioning

confidence: 99%

Quantitative Structure Activity Relationship Prediction of Oral Bioavailabilities Using Support Vector Machine

Fatemi¹,

Fadaei

2014

Journal of the Korean Chemical Society

View full text Add to dashboard Cite

ABSTRACT. A quantitative structure activity relationship (QSAR) study is performed for modeling and prediction of oral bioavailabilities of 216 diverse set of drugs. After calculation and screening of molecular descriptors, linear and nonlinear models were developed by using multiple linear regression (MLR), artificial neural network (ANN), support vector machine (SVM) and random forest (RF) techniques. Comparison between statistical parameters of these models indicates the suitability of SVM over other models. The root mean square errors of SVM model were 5.933 and 4.934 for training and test sets, respectively. Robustness and reliability of the developed SVM model was evaluated by performing of leave many out cross validation test, which produces the statistic of Q SVM 2 = 0.603 and SPRESS = 7.902. Moreover, the chemical applicability domains of model were determined via leverage approach. The results of this study revealed the applicability of QSAR approach by using SVM in prediction of oral bioavailability of drugs.

show abstract

Untitled

Cited by 130 publications

References 4 publications

Identification of novel cannabinoid receptor ligands via evolutionary de novo design and rapid parallel synthesis

Identification of novel cannabinoid receptor ligands via evolutionary de novo design and rapid parallel synthesis

Prediction models of human plasma protein binding rate and oral bioavailability derived by using GA–CG–SVM method

Quantitative Structure Activity Relationship Prediction of Oral Bioavailabilities Using Support Vector Machine

Contact Info

Product

Resources

About