scispace - formally typeset
Search or ask a question

Showing papers by "Georgia Melagraki published in 2013"


Journal ArticleDOI
TL;DR: In this article, a database of relative binding affinity of a large number of estrogen receptor (ER) and/or ERβ ligands was assembled (546 for ERα and 137 for ERβ) and both single-task learning (STL) and multitask learning (MTL) continuous quantitative structure-activity relationship (QSAR) models were developed for predicting ligand binding affinity to ERα or ERβ.

75 citations


Journal ArticleDOI
TL;DR: This work introduces the custom made Enalos KNIME nodes, made publicly available by Novamechanics Ltd, as key nodes to develop robust and validated quantitative structure–property models (QSPRs).

73 citations


Journal ArticleDOI
TL;DR: Docking studies of the representative highly active 12b clearly showed that this molecule has an extra hydrophobic binding feature compared to prototype drug Losartan and it fits to the extra hydphobic cavity, which may contribute to the discovery and development of a new class of biologically active molecules through bis-alkylation of the imidazole ring by a convenient and cost effective synthetic strategy.

27 citations


Journal ArticleDOI
TL;DR: The proposed method, due to the high predictive ability, can be a useful aid to the design of new high-quality products with minimum investment in time and money for experimental work.
Abstract: Environmental durability, ageing and corrosion tests are applied before a product comes to the market to ensure that the product's life time behaviour is well established. Current tests demand a long period of weeks - months before safe results can be concluded based on the procedures described and implemented so far. The need for a reduction in time required for evaluation of a product's behaviour is now more than important since now there is a limited time from development until the product reaches the market. The pressure of the market, the time gap between the conception of a product and its production decreases dramatically and there is a need for better and more severe corrosion and ageing tests in order to shorten the lead-time of products. To achieve this goal, in an effort to reduce the cost and time required, within EUREKA MAAC project [1], we have computationally explored experimental data produced by our partners and from the literature. We have developed advanced mathematical models for predicting environmental durability, ageing resistance and corrosion inhibition with respect to environmental factors and material structure. For this purpose we have successfully developed KNIME workflows including Enalos KNIME nodes [2] such as Enalos Mold2 KNIME node, Enalos Model Acceptability Criteria KNIME Node and Enalos Domain KNIME Node that were implemented to calculate Mold2 descriptors, validate the produced models and define its domain of applicability respectively [3]. The proposed method, due to the high predictive ability, can be a useful aid to the design of new high-quality products with minimum investment in time and money for experimental work.