Due to the rapid development of information technology, in silico methods are used in numerous branches of science, next to in vitro and in vivo solutions, contributing to the success of researches. Instead of traditional empirical experiments, these methods use computer simulations, calculations, which can significantly reduce costs and time commitment whilst keeping the results trustworthy. Thus, it is not surprising, that recently, this form of experiment has become utterly popular, however, in order to accomplish authentic results, these algorithms need to be optimized and improved. There are many on-going efforts to reach this goal, loads of new methods are introduced continuously.
The present thesis gives a picture of the role of in silico methods in pharmacology and explains the used algorithms in general and thoroughly examines their individual and ensemble performance. It also enhances the chance to get a more effective solution by aggregating the algorithms’ predictive ability.
The first crucial part of the drug discovery process is identifying targets and determining the molecules, chemical compounds that will affect them. The hereby introduced in silico procedures play a significant role in this process, their continuous improvement is starting to be a vital part of pharmacology. It means a long term help in drug discovery.
It is worth mentioning, that the solutions detailed in the current document can be useful not only in the field of pharmacology, but also in case of other problems requiring predictive processes (e.g. book-, movie- and music recommending systems).