When studying large biomolecules or long-timescale effects, the computational cost of all-atom simulations becomes huge, so such studies are in practice unfeasible. This difficulty is avoided by proposing “coarse-grained” models of the biomolecules, moving thus to a mesoscopic level of description.
The Peyrard-Bishop-Dauxois (PBD) model of DNA gives an adequate description of some of the physical properties of the DNA double strand, as the melting transition or bubble formation. We have used such a model to study the mechanical denaturation [Cuesta05] and the formation of bubbles in promoter sequences [Tapia10]. The relation between the location of such bubbles in the sequence and the sites of biological relevance (as protein binding-sites) is a question of great biological interest.
This latter study has led us reciently to extend PBD model by introducing the interaction of the DNA chain with a test-particle [Tapia12]. This test-particle is inspired in DNA-binding proteins, as it explores the DNA sequence by diffusing along the chaina and interacting with it in an especific way. This new model allows the identification of biologically active sites. Indeed, by applying a numerical method developed in our group [Prada10], we are able to map the molecular dynamics trajectories onto a Conformational Markov Network and build the free-energy landscape of the DNA promoter, extracting relevant information about such sites.
One the other hand, proteins have become one of the most sudied systems in biophysics, due to the important role they play in most biological processes at the molecular level. From the “coarse-grained” perspective, different approaches exist. One of our interests are Gaussian-Network models. These models start with the positions of the alpha-carbons in the native structure and build a network by considering in contact those closer than a certain threshold. This model can be extended by allowing contacts to be broken through a phenomenological potential. This consideration allows the study of dynamical properties such as the denaturation transition. The problem of protein folding can also be attacked from the “coarse-grained” perspective. HP models clasify the aminoacids according to their behavior with respect to water, introducing thus different interactions among them that allow to mimic the complex dynamics of protein folding.
Publications:
[Cuesta05] S.Cuesta-López, J. Errami, F. Falo, M. Peyrard. Can we model DNA at the mesoscale? Journal of Biological Physics 31 273-301 (2005).
[Tapia10] R. Tapia-Rojo, J. J. Mazo, F. Falo. Thermal and mechanical properties of a DNA model with solvation barrier. Physical Review E, 82, 031916 (2010)
[Tapia12] R. Tapia-Rojo, D. Prada-Gracia, J. J. Mazo, F. Falo. Mesoscopic Model for Free Energy Landscape Analysis of DNA sequences. Physical Review E, 86, 021908 (2012)
[Prada10] D. Prada-Gracia, J. Gómez-Gardeñes, P. Echenique, F. Falo. Exploring the free energy landscape: From dynamics to networks and back. PLoS Computational Biology 5, e1000415. (2009).
