Relationship between intraocular pressure lowering effect and chemical structure of imidazo[1,2-a]benzimidazole and pyrimido[1,2-a]benzimidazole derivatives

This article contains data that relate to the study carried out in the work of Marcus et al. (2018) [1]. Data represent an information about pharmacophore analysis of imidazo[1,2-a]benzimidazole and pyrimido[1,2-a]benzimidazole derivatives and results of construction of the relationship between intraocular pressure (IOP) lowering activity and hypotensive activity of imidazo[1,2-a]benzimidazole and pyrimido[1,2-a]benzimidazole derivatives using a multilayer perceptron artificial neural network. In particular, they include the ones listed in this article: 1) table of all pharmacophores of imidazo[1,2-a]benzimidazole and pyrimido[1,2-a]benzimidazole derivatives that showed IOP lowering activity; 2) table of all pharmacophores of the compounds that showed absence of IOP lowering activity; 3) table of initial data for artificial neural network analysis of relationship between IOP activity and hypotensive activity of this chemical series; 4) graphical representation of the best neural network model of this dependence; 5) original txt-file of results of pharmacophore analysis; 6) xls-file of initial data for neural network modeling; 7) original stw-file of results of neural network modeling; 8) original xml-file of the best neural network model of dependence between IOP lowering activity and hypotensive activity of these azole derivatives. The data may be useful for researchers interested in designing new drug substances and will contribute to understanding of the mechanisms of IOP lowering activity.

model of dependence between IOP lowering activity and hypotensive activity of these azole derivatives. The data may be useful for researchers interested in designing new drug substances and will contribute to understanding of the mechanisms of IOP lowering activity.
& 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Subject area
Medicine More specific subject area Pharmacology, QSAR, In silico drug design Type of data Tables, figure, *.txt, *.xls, *.stw and *.xml files How data was acquired This study was done in ocular normotensive rats and rebound tonometry (Tonolab, Icare Finland) was used to estimate intraocular pressure (IOP). Pharmacophore analysis was carried out using IT Microcosm package (Russian Federation). Neural network modeling was performed using Statistica 6.0 package (StatSoft Inc., USA).

Data format Analyzed Experimental factors
This data is supplementary to article [1]. A total of 27 new compounds were synthesized as described previously and tested for IOP lowering effect in rats. These compounds included twenty 9H-imidazo [ The data are available in this article and in appended files.

Related research article
This data is supplementary to article [1].

Value of the Data
The data include the results of pharmacophore analysis of IOP lowering activity of imidazo [1,2-a] benzimidazole and pyrimido[1,2-a]benzimidazole derivatives and may be useful for researchers interested in designing new drug substances.
IOP lowering activity and pharmacophore list will help other researchers in investigating new drugs. These data can be compared with the data of pharmacophore analysis performed by other researchers and this will facilitate international collaborations in the field of drug development.
The results of analysis of the relationship between IOP activity and hypotensive activity of imidazo [1,2-a]benzimidazole and pyrimido[1,2-a]benzimidazole derivatives with the help of artificial neural networks will contribute to understanding of the mechanisms of IOP lowering activity.
The animal studies were done in compliance with the ARVO statement for use of animals for vision research and the institutional ethical guidelines. To evaluate 27 imidazobenzimidazoles derivatives for their IOP lowering effect, 3 different concentrations 0.1%, 0.2% and 0.4% were prepared for topical   weighed to 0.4 mg and dissolved in 1 ml (0.25%) HPMC, then serial dilution was done to obtain 0.2% and 0.1% concentrations. The remaining water insoluble compound was dissolved in 0.1% DMSO in 0.25% HPMC and similarly 3 concentrations of this compound were prepared for topical application. IOP was measured in the conscious rats using TonoLab (Icare, Finland) rebound tonometer specifically designed for rodents (rat/mouse). Since it is a noncontact tonometer, it does not require use of an anaesthetic agent. The TonoLab was placed right at the centre of the cornea and the distance from the tip of the probe to surface of the cornea was 1-4 mm. For this study, 3 rats were used in each group and the left eye (TE) served as treatment eye while the right eye served as control eye (CE). IOP was measured at 0.5, 1, 1.5, 2, 3, 4, 5 and 6 h post-instillation. Six readings were obtained at each time point and the mean was taken as the final measurement.
Hypotensive activity of tested compounds was evaluated in anesthetized animals (pentobarbital, i.p. 50 mg/kg/bw, JSC Tallinn Pharmaceutical Plant, Estonia) as described previously [12,13]. Tested compounds were administered in jugular vein. Systemic arterial pressure (SAP) was recorded through carotid artery for 1 h after the administration of the compound using a mercury manometer ПМР-2 (Russian Federation). The measure of hypotensive activity was presented as ED 20 , a concentration (mol/kg) causing maximum SAP to decrease by 20% in 1 h. Additionally, scale was introduced to assess the potency of compounds [14]: index (Ind) of 3 points was applied for ED 20 r 4.0 mg/kg, 2 points for ED 20 The separation of the studied substances into active and inactive classes was carried out by means of a cluster analysis of 6 indicators of IOP lowering activity by the k-means method using the Statistica 6.0 package [15].
The pharmacophore analysis was performed using the IT Microcosm 7.2 package [14]. First, the chemical structures were translated into descriptors of the QL language [16] using the utilities ActUtil, TranQL2, and MakeData. Then, with the help of the FarmFor module, for each type of QL descriptors, the following were calculated: P afrequency for a class of active compounds; P ifrequency for a class of inactive compounds; Prsignificance in hypergeometric test [16]. The QL descriptor was considered a potential pharmacophore of presence of IOP lowering activity if Pr r 0.05 and P a 4 P i . The QL descriptor was considered a potential pharmacophore of absence of IOP lowering activity if Pr r 0.05 and P i 4 P a .
Neural network simulation of relationship between IOP and hypotensive activities was performed by method of multi-layer perceptron artificial neural networks using the Statistica 6.0 package [15]. Within the framework of the classification model, the architecture of the neural network in the form of a two-layer perceptron was used. As input neurons there were four indicators of hypotensive activity: ED 20 (mg/kg), Ind, ED 20 (μM/kg), Lev (the meaning of these parameters is described in [14]). As output neurons, IOP lowering activity indicators were: IOP Cluster (1)presence of activity, IOP Cluster (0)absence of activity. Neural networks were constructed in automatic mode ANN, random sampling 80% for training and 20% for testing. The following parameters of the neural network simulation were set: 1) multilayer perceptron MLP; 2) the minimum number of hidden neurons 3; 3) the maximum number of hidden neurons 10; 4) number of trained networks 1000; 5) number of selected good networks 50; 5) types of activation functions for hidden and output neurons Identity, Logistic, Tanh, Exponential, Sine, Softmax; 7) other parameters were accepted by default. After training, from the 50 automatically selected good networks, based on the accuracy of training and testing, the best network was selected. The statistical correspondence of the experimental and calculated estimates of IOP lowering activity was determined using Statistica 6.0 package [15] by means of nonparametric Spearman correlation coefficient.