Elucidating molecular mechanisms of biomolecule functions is crucial for advancing understanding of biological processes, pathology of many hereditary and acquired diseases, as well as developing novel therapeutic strategies or targeting drugs. Recent years have witnessed a huge pike on high resolution structures of biologically critical membrane proteins shedding lights on atomistic level details, meanwhile functional studies have provided us with abundant macroscopic functional measurements upon perturbations, such as mutations and change of experimental conditions. Rationalizing those functional data in the molecular level based on resolved structures has been an important but challenging knowledge gap, which can be bridged by dynamics studies through computational simulations achieving high spatial and temporal resolutions. In Chapter two to four, we focused on studying gating and regulation mechanisms of two membrane channel protein: the TMEM16F lipid scramblase and the TRPV4 ion channel. We performed all-atom simulations on the inner gate charged TMEM16F mutants and observed spontaneous the pore opening and hydration process. The sampled closed-to-open transition detailed in pore-forming helices movements on different specific directions in the atomistic level. The resulting open state was found to be functionally active since a direct lipid scrambling event was sampled. Our study predicted a possible biologically relevant open state eluded from previous cryo-EM studies and revealed the activation transition process as well as the ion and lipid translocation pathways. In the TRPV4 gating study, we aimed at resolving free energy barrier contributions from the classic bundle-crossing gating mechanism and the hydrophobic gating mechanism. We first demonstrated the dewetting transition can readily happen in the bundle-crossing and highly hydrophobic pore. Through thorough free energy calculations, we showed that disruption of the hydrophobicity of the pore can significantly reduce the ion permeation barrier through lowering down the hydration free energies of the lower pore region. In Chapter four, we presented a story of PI(4,5)P2 regulation of the TRPV4 ion channel. We first disapproved previously proposed PI(4,5)P2 binding sites and identified two more plausible binding sites in the inner leaflet-protomer interface. Relative free energy calculations between the two sites showed they almost have similar binding affinities, indicating the possibility of dynamic binding of multiple PI(4,5)P2 for regulation. Careful dynamic coupling network analysis further resolved dynamics implications of binding in those two sites, suggesting PI(4,5)P2 can prime the channel for temperature activation. In Chapter five, we used extensive coarse-grained simulations to examine important domain-domain interaction modes in the chemotaxis protein CheA, the autokinase initiating phosphorylation cascade to regulate downstream flagellar rotation for motile bacteria to swim towards favorable environment. Important P1-P4 trans- productive contacting modes and P1/P1’ dimerization modes were predicted, generating testable measurements and hypotheses for experimental validations. All Chapters together showed how powerful it is to combine functional studies, structural biology and computational simulations for investigating functional and regulatory mechanisms of biomolecules in the biologically relevant environment.