Development of a model of metabolic pathways in Python for testing Sildenafil on smooth muscle and molecular docking

Abstract

In recent decades, bioengineering has developed a new approach to biology, aimed not only at the realization of analytical and diagnostic tools, but at the understanding and simulation of the molecular phenomena that take place in cells. This has resulted in a rapid spread of “in silico” modeling of biological processes around the world.

The expression “in silico” refers to the use of computers to create models necessary for carrying out biological studies. Models of this type have several advantages: the reduction of costs for companies and research laboratories, the significant saving of time thanks to the instantaneous predictions, the zeroing or reduction of the number of laboratory animals used and the possibility to use user-friendly graphical interfaces, which allow even novice users to enter the world of “in silico” modeling. On the other hand, their main limitations must also be specified: the scarce availability of data, the simplifications adopted and the high number of interconnections present between the various cellular molecular phenomena.

The aim of this thesis work is the development of a model in Python able to simulate metabolic pathways to describe how the drug of interest, previously selected upstream, interacts and produces the desired effects. Specifically, first of all the metabolic pathway that leads to smooth muscle relaxation was modeled and then the effects of the drug Viagra, containing the active ingredient Sildenafil, were modeled on the same metabolic pathway, in order to verify its compatibility for clinical applications other than those for which the drug is commonly used.

It was decided to use Python since it is a high-level programming language, object-oriented and widespread throughout the scientific community for its advantages: free and free use on many operating systems, support for functional programming and the rich ecosystem of searchable mathematical and scientific libraries, such as NumPy, SciPy, and PySB. In addition, Python is able to interface with other programming languages, such as C and C ++, playing the role of glue for applications that use libraries written with the aforementioned languages. Obviously, there are other languages developed specifically for biological modeling that use a specialized syntax to code the processes of biological models. However, these languages lack many features present in Python that can be used to make complex code more human readable, such as conditional programming constructs, loops, functions, classes, and modules.

In addition, the work presents a molecular docking study between Sildenafil and the receptor protein, through the AutoDock Vina software to evaluate the interaction between the active ingredient and the biological structure.

Molecular docking is a computational method that attempts to efficiently predict the interaction between two molecules or, more frequently, between a macromolecule (receptor) and a small molecule (ligand), starting from their unbound structures. AutoDock Vina is an open-source program for doing molecular docking. It was chosen for its computational potential, in fact, it has significant optimizations in terms of speed and accuracy in predicting the link, compared to the predecessor software AutoDock.

Objectives

The purpose of this thesis work is the development of a model in Python able to simulate metabolic pathways to describe how the drug, previously selected upstream, interacts and produces the desired effects. Specifically, the metabolic pathway that leads to smooth muscle relaxation was first modeled and then, the effects of the drug Viagra, containing the active ingredient Sildenafil, were modeled on the same metabolic pathway, in order to verify its compatibility for clinical applications other than those for which the drug is commonly used.

Research methodology

Python modeling and molecular docking simulations

Conclusions

The work is a model to simulate the metabolism and interactions of the active ingredient Sildenafil, both in approved clinical applications, i.e. the treatment of erectile dysfunction and pulmonary hypertension, and in new, not yet approved clinical applications. From the results obtained, we can see that the model developed is able to predict, quite accurately, the effect of Sildenafil on smooth muscle and its interactions with the PDE5 protein. In fact, as predicted, it has been shown that the temporal course of cGMP in the presence of Sildenafil is reduced more slowly, resulting in a more decisive and longer smooth muscle relaxation effect. In this way, it is possible to provide information also for future simulations with concentrations different from those used in the model. While, as far as the interactions are concerned, the configuration calculated by the AutoDock Vina software has minimal differences with the actual laying data present in the literature, in order to have important information for the possible docking of Sildenafil with other molecules. In the literature, there is no modeling in Python of the following metabolic pathway and of the action of Sildenafil, therefore, from this point of view, this model represents an innovation. In fact, information was found regarding the pathway reactions considered individually, but there is no modeling in which the entire cascade of reactions is considered as a whole. Simulating the reactions within the metabolic network provides a greater similarity with reality, rather than considering them separately.

Future developments

The various functions defined in the model can be used for the analysis of the metabolism of other drugs, as they have been defined in a general way and not as a function of this particular pathway. The functions have been conceived as user-friendly, that is even less experienced users in programming may be able to model the chemical reactions of interest, as long as they have all the necessary parameters. The development of this type of model involves a significant reduction in costs, times and the number of in vivo experiments on animals. Therefore, the model is versatile for numerous applications: as a reference model for the implementation of other metabolic pathways, to perform an initial screening of target molecules in order to reduce their number or to validate experimental protocols.