• accelerated molecular dynamics
• excitonic model
• homology modeling
• LHCSR
• light harvesting
• light harvesting complexes stress related
• molecular dynamics
• mustiscale polarizable QM/MM
• Physcomitrella patens
• pigment protein complex
• polarizable embedding

Light Harvesting Complexes (LHCs) are a superfamily of pigment-protein complexes, containing chlorophylls and carotenoids, responsible for photon absorption and transfer of the excitation energy to the reaction center in photosystems. A sub-family of LHCs, Light Harvesting complexes stress-related, is expressed in stress-related conditions and is responsible for quenching excess energy in the photosystem. At variance with other LHCs, there is no information on the three-dimensional structure of any protein in the LHCSR family.

In this thesis we present an approach for structure modeling and refinement of LHCSR1 of the Physcomitrella patens moss, which shares only ~30% sequence similarity to the other LHC complexes. Homology modeling was complemented by classical molecular dynamics (MD) simulations to refine the structure and to identify the stable parts of the complex, the transmembrane helices, and the highly mobile loop regions and unfolding helices. The structure was further refined with the enhanced sampling technique accelerated MD. Multiscale polarizable QM/MM calculations of the excitonic absorption and circular dichroism spectra were used to validate the obtained structure against experiments.

The protocol proposed in this thesis proved to be effective in generating a realistic model of LHCSR1 and it promises to be a viable strategy for the modeling of LHC proteins for which a high-resolution structure is not known experimentally.