The interaction between proteins and solid surfaces is of fundamental importance for a variety of scientific and technological problems, ranging from biomineralization to nanoelectronics . Combinatorial biotechniques have shown that the molecular recognition capability of proteins can be specifically oriented toward inorganic surfaces, opening-up the way for new concepts in self-assembling . The interaction of proteins with nanostructured surfaces or nanoparticles is also the key to several biological processes occurring in vitro and in vivo . However, to date the demonstration of the importance of protein-surface interactions has not been accompanied by an understanding of the mechanisms that determine the partnership and the resulting function. What features of the surface and of the proteins determine which protein is able to bind to a given surface and how? Are electronic, structural, morphological or dynamic factors most relevant? Is the protein structure modified by the interaction with the surface, and how?
Experiments in this field are extremely challenging (they need to probe the interface between the biomolecule and the solid interface in realistic aqueous environment) and cannot provide the level of structural and energetics details needed for a molecular understanding of protein-surface systems. Computational approaches (in particular atomistic simulations) have the potentials for unraveling the microscopic picture behind protein-surface interactions. However, these systems are challenging for computations. First, they span several length scales (from the local interaction of side chain to the extension of entire proteins and nanoparticles) and several time scales (from the few ps of local rearrangements to the ms, s if not hours necessary for major conformational changes of proteins or deposition of protein layers). Therefore, the development of multiscale approaches apt to treat hybrid protein-inorganic systems is a fundamental computational target. Moreover, proteins and solid surfaces or nanoparticles have been traditionally studied by different scientific communities, and the computational tools developed for one kind of systems (eg, proteins) are not often compatible with the tools developed for the others. Of course methods with a minimal number of assumptions (such as ab initio methods) can already treat both systems together (although implementation details such as plane-wave vs localized basis sets may be more suitable for a system or the other), but the high computational cost of such methods do not allow to study systems larger than hundreds-one thousand of atoms for more than a few tens of ps .
Since the pioneering paper by Braun et al.  (who performed classical atomistic MD simulations of a gold binding peptides on two different crystal surfaces of gold), there have been in the literature scattered attempts to study protein-surface interactions, mostly with mesoscopic models (see the review by J.J. Gray). More recently, the focus has been again on classical MD simulations, the single-scale method that presently better compromises between accuracy and size of simulated systems and time length of the simulations. Ab initio methods have been used to treat the adsorption of amino acids on inorganic surfaces . Recently, ab initio MD simulations of model proteins on inorganic surfaces in water have been also performed and, although the size of the system and the length of the simulation are limited by the high computational costs, useful information on the protein-surface interactions and the role of water could be obtained . Ab initio calculations have been also used to parametrize classical force fields for amino acid - surface interactions, that have been then exploit to perform MD simulations [8-10].
In one case, computational tools have already been demonstrated to have predicting power . However, general strategies to build computational tools for simulating and predicting protein-surface interactions are not yet available.
The current challenges in the field of simulating and modeling protein surface interactions are summarized in :
• Protein-surface systems cover a huge diversity which poses a challenge for the derivation of accurate, consistent force fields. One needs to consider the properties of the surface material and the protein, as well as the aqueous solution in which they are immersed. The level of details necessary to obtain reliable results depends on the properties under study. Force field evaluation or modification needs experimental benchmark data sets, that should not only include energetic data, but also structural information
• Extensive sampling of conformations is another fundamental problem to be addressed for gaining insights into the binding process. In fact, a proper sampling of the conformational space is not only needed to properly account for entropic contributions to binding, but also to account for the different protein conformations that may exist in solution, and interact with the surface.
• A key issue in the context of simulations is bridging time and length scales. Protein-surface systems are in fact spanning length scale from the atomics (local interactions between amino acids and the surface) to the nm of even microm one, if one considers the building of protein layers on surfaces or nanoparticles. Moreover, rearrangement of protein conformations (induced by the surface) may take place on the same long time scale (ms, s) that characterizes protein rearrangement in solution. For this reasons, the development of multiscale methods for these systems is an important task.
• The role of water and ions in the protein-surface interaction process is extremely important, and thus they require an accurate description. Some MD simulations with explicit water show that, for strongly polar surfaces, binding is not to the surface itself, but rather to a structured water layer on the surface . Even when the binding is with the surface, the structuring of water close to the surface is not negligible and, often, continuum solvation models for bulk water cannot be applied . Not only water, but also ions play a fundamental role in controlling interactions at interfaces. In fact, ion concentration may even reverse the sign of the interaction between charged surfaces.
Finally, we should add to the list the complexity of real inorganic systems used in applications. Even the more well-behaving real surfaces have defects (adatoms, vacancies, steps, etc.), not to mention the transformations that inorganic surfaces undergo when in contact with liquid water or with the protein. Such defects may be highly reactive and contribute substantially to the surface behavior even if their population is small. Nanocrystals also present a large variety of interactions sites (faces, edges, vertexes); they can also be covered with surfactants that alter their interaction with water and proteins.
All these situations represent yet unexplored variables in the field of protein-inorganic surface and nanoparticle interactions, that require further extension of computational tools (eg., development of proper force fields) and methodologies.