The main goal of research presented in this work is to decipher mechanisms of processesin macromolecular biological systems. Several such case studies are presented in this thesis. In biological systems a majority of sub-molecular atomistic mechanisms of functionally important processes are not accessible by experimental approaches. However, modern methods of computational chemistry and molecular modeling are suitable with the right application to step in and study the molecular mechanisms of some of these phenomena. The main approach used in this work is, for a given problem, to pick and match an appropriate combination of computational methods out of a modern theoretical tool kit. The first problem presented in this thesis is to predict pKas of amino acid residues in proteins. We have implemented an efficient algorithm to apply a Free Energy Perturbation (FEP) method to calculation of pKas. We have then established a link of pKas of single amino acid residues with global conformational rearrangements of the protein. The second problem presented here is to model a pathway by which a bacterial scavenger protein extracts an iron heme group from a hemoglobin protein. The simulations allowed us to elucidate a mechanism of a heme extraction and to uncover driving forces for such extraction, as well as to identify key amino acid residues involved in the heme transfer. The third problem presented in this work is to develop a mechanism of a divalent ion permeation through a calcium selective epithelial ion channel TRPV6. We have also studied a difference in permeation mechanisms of different mono- and divalent cations through the TRPV6 channel. The forth problem presented in this thesis is conformational analysis of a backbone branched RNA (bbRNA) in free and bound to a protein states. The final problem described in this work is to understand mechanism of interactions of non-competitive inhibitors with a glutamate receptor.