Studies in some Aspects of Anti-Tubercular Drug Design Utilizing Theoretical Molecular Descriptors
EPrints@IICB
View Archive InfoField | Value | |
Title |
Studies in some Aspects of Anti-Tubercular Drug Design Utilizing Theoretical Molecular Descriptors
|
|
Creator |
Ghosh, Payel
|
|
Subject |
Structural Biology & Bioinformatics
|
|
Description |
Rational drug designing adopts an integrated approach to study and understand the structures of potential agents (particularly, the analogs of an existing drug or lead compound), elucidate how they might interact with target protein structures and predict their activity profiles in-silico. Identification of such potential agents from a large number of analogous structures makes computational modeling approach a useful tool for drug discovery studies. Availability of various experimental data against different leads and their derivatives, and sometimes, the structural knowledge of their target proteins facilitate the structure-activity modeling studies and make modern day approach to drug design “rational”, compared to the protocols pursued in the past. Quantitative structureactivity relationship (QSAR) can be derived for a series of structurally similar type of derivatives, and efficient QSAR models can predict the activity profiles for any structurally known molecule in that series, whether real or virtual. Prediction of such important properties for a large number of analogous structures definitely would be a “blessing” for drug discovery process and will reduce the cost and time of the entire drug development process. Synthetic chemists can utilize these models as a decision support tool in synthesis planning and do the initial screening in a fraction of time. Molecular docking, another important chemometric technique, has the potential to address several important issues that arise in drug discovery such as interaction between the ligand (chemical compound) and the target (generally, protein structure). The thesis enlists various QSAR models in conjunction with molecular docking studies for mainly three types of chemical structures having potential anti-tubercular activity, viz., derivatives of fluoroquinolone, quinoxaline and nitrofuranyl amide and describes the usefulness of different statistical as well as neural network approaches to improve the predictability of these models. Insights gained from these studies may be applied in anti-tubercular drug design with additional chemical structures. Mycobacterium tuberculosis (Mtb), the principal causative agent of tuberculosis in humans, is estimated to cause two million deaths every year. The existing drugs viz., isoniazid, rifampicin, ethambutol, pyrazinamide etc., although of immense value in controlling the disease to some extent, have several shortcomings, the most important of them being the emergence of drug resistance rendering even the front-line drugs inactive. So, there is an urgent need to develop novel agents having high anti-tubercular activity values, which may shorten the treatment period of tuberculosis disease, thus preventing the emergence of drug resistance.The thesis is mainly concerned with the development of novel computational approaches that could be useful in the early stages of drug discovery for elucidating the potency of promising anti-tubercular agents belonging to several different families of anti-tuberculosis leads. The thesis begins with a focus on the fluoroquinolone derivatives, which are used in case of first line drug resistance. To examine specific structure activity relationships of quinolone antibacterials against mycobacterial activity, an attempt has been made to establish a quantitative structure activity relationship modeling for a series of quinolone compounds against Mycobacterium fortuitum and Mycobacterium smegmatis. Due to lack of sufficient physicochemical data for the anti-mycobacterial compounds, it becomes very difficult to develop predictive methods based on experimental data. QSARs were generated from the standpoint of physicochemical, constitutional, geometrical, electrostatic and topological indices, which have been calculated solely from the chemical structure of N-1, C-7 and 8 substituted quinolone compounds and ridge regression models have been developed for a better understanding of structure-activity relationships. Consideration of an intermolecular similarity analysis approach based on atom-pair method led to data sub-grouping of the considered analog structures depending on their similarity to a drug or lead molecule, and the influence of various molecular descriptors in different data subsets were compared. It is suggested that topological descriptors play an important role in the design of QSAR models for fluoroquinolone derivatives. In continuation of the work with fluoroquinolone derivatives, the applicability and scope of descriptor based QSAR models have been improved by introducing feature selection method along with the modeling techniques. Virtual screening using molecular docking approach have also been applied in course of this work to identify potential molecules targeting DNA gyrase A from Mycobacterium tuberculosis, an effective and validated mycobacterial target. Initially QSAR models have been developed against M.fortuitum and M. smegmatis using series of structurally related fluoroquinolone derivatives as DNA gyrase inhibitors. The statistically significant models have been then validated by a test set of compounds and y-randomization scheme. To aid the creation of novel antitubercular compounds, combinatorial library has been developed on fluoroquinolone template whose activity values have been measured by the above models. Highly active compounds predicted from the models were then subjected to molecular docking study to investigate the mechanism of drug binding with the DNA gyrase A protein of M. tuberculosis and the compounds showing similar type of binding patterns with that of the existing drug molecules viz., sparfloxacin, have been reported. Observations suggest that hydrophobic characteristics of the small molecular structure together with few hydrogen bond interactions are playing an essential role in anti-tubercular activity for the fluoroquinolone derivatives. A representative set of seven compounds with highly predicted MIC values were identified in the analysis. An endeavor towards development of QSAR models for quinoxaline compounds having excellent anti-tubercular activities constitute another integral part of the investigation. Since, there is a dearth of anti-mycobacterial activity studies involving an adequate number of quinoxaline molecules, an effort was made to merge multiple quinoxaline data sets after verifying that their constituents share the same chemical space.Such a merged data set was utilized to develop robust QSAR models that were validated statistically by leave-one-out and leave-many-out methods. Both 2D and 3D-QSAR models have been developed for the quinoxaline derivatives. In addition, genetic algorithm (GA) and simulated annealing (SA) have been applied as variable selection methods for the selection of a preferred set of molecular descriptors that can signify the chemico–biological interaction. 2D-QSAR modeling using GA or SA based partial least squares (GA-PLS and SA-PLS) methods identified some important topological and electrostatic descriptors as vital factors for tubercular activity. Kohonen network and counter propagation neural network (CP-NN) considering GA and SA based feature selection methods have also been applied in an attempt to capture inherent non-linearity in the structure-function relationships of quinoxaline compounds. In addition, 3D-QSAR models obtained by GA-PLS and SA-PLS methods identify the influences of steric and electrostatic field effects that can give a direction for the synthesis of new quinoxaline derivatives with potent anti-tubercular activity. A series of nitrofuranyl amides were also subjected to quantitative structureactivity relationship (QSAR) analysis using various feature selection methods. Genetic algorithm (GA), simulated annealing (SA) and stepwise regression have been applied as variable selection methods for an effective comparison and subsequently, models were generated for these compounds. Both 2D and 3D QSAR analyses of such derivatives provide important structural insights for designing potent anti-tuberculosis drugs.Summarising, the thesis presents a comprehensive in silico analysis of structureactivity relationships of three important classes of drugs and their derivatives having antitubercular activity. Although the principal focus of the thesis has been the discovery of potent anti-tubercular compounds, the methodologies described are highly generic in nature, lending themselves to the application in drug discovery programmes targeted against other diseases. |
|
Date |
2010
|
|
Type |
Thesis
NonPeerReviewed |
|
Format |
application/pdf
|
|
Identifier |
http://www.eprints.iicb.res.in/1877/1/PAYEL_GHOSH_PHD_THESIS.pdf
Ghosh, Payel (2010) Studies in some Aspects of Anti-Tubercular Drug Design Utilizing Theoretical Molecular Descriptors. PhD thesis, Jadavpur University. |
|
Relation |
http://www.eprints.iicb.res.in/1877/
|
|