Global QSAR modeling of LogP values of phenethylamines acting as adrenergic alpha-1 receptor agonists

• Published in: Interdisciplinary sciences : computational life sciences Vol. 5; no. 2; pp. 150 - 154
• Main Authors: Yadav, Mukesh, Joshi, Shobha, Nayarisseri, Anuraj, Jain, Anuja, Hussain, Aabid, Dubey, Tushar
• Format: Journal Article
• Language: English
• Published: Springer International Association of Scientists in the Interdisciplinary Areas 01.06.2013; Springer Nature B.V
• Subjects: Adrenergic alpha-1 Receptor Agonists - chemistry; Adrenergics; Atomic properties; Biological; Biomedical and Life Sciences; Computational Biology/Bioinformatics; Computational Science and Engineering; Computer Appl. in Life Sciences; Databases, Chemical; Derivatives; Health Sciences; Life Sciences; Mathematical and Computational Physics; Mathematical models; Medicine; Models, Molecular; Molecular structure; Phenethylamines - chemistry; Physical properties; Quantitative Structure-Activity Relationship; Reproducibility of Results; Statistics as Topic; Statistics for Life Sciences; Theoretical; Theoretical and Computational Chemistry; Toxicity; adrenergic alpha-1 receptor agonists; logP; phenethylamines; global QSAR
• ISSN: 1913-2751; 1867-1462
• DOI: 10.1007/s12539-013-0162-0

Global QSAR models predict biological response of molecular structures which are generic in particular class. A global QSAR dataset admits structural features derived from larger chemical space, intricate to model but more applicable in medicinal chemistry. The present work is global in either sense of structural diversity in QSAR dataset or large number of descriptor input. Forty phenethylamine structure derivatives were selected from a large pool (904) of similar phenethylamines available in Pubchem database. LogP values of selected candidates were collected from physical properties database (PHYSPROP) determined in identical set of conditions. Attempts to model logP value have produced significant QSAR models. MLR aided linear one-variable and two-variable QSAR models with their respective R 2 (0.866, 0.937), R 2 A (0.862, 0.932), F-stat (181.936, 199.812) and Standard Error (0.365, 0.255) are statistically fit and found predictive after internal validation and external validation. The descriptors chosen after improvisation and optimization reveal mechanistic part of work in terms of Verhaar model of Fish base-line toxicity from MLOGP, i.e . (BLTF96) and 3D-MoRSE -signal 15 /unweighted molecular descriptor calculated by summing atom weights viewed by a different angular scattering function (Mor15u) are crucial in regulation of logP values of phenethylamines.