Computational Biophysics - Prof. Paolo Carloni

Our lab combines different computational molecular biology approaches in a powerful strategy aimed at dissecting structural and energetic facets in cellular pathways related to perception and deranged cascades of events in molecular medicine. A major effort is devoted to the inclusion in the calculations of the effects due to the cellular environment. We also enjoy collaborations with several theoretical groups. Because of the tremendous complexity of the systems under investigation, comparison with experiment is absolutely required to test any hypothesis arising from the calculations. In fact, almost all of our work is in collaboration with molecular and cell biologists and experimental biophysicists. We compare our results against both in vitro and in vivo essays.

Molecular recognition in biology

The interactions among biological macromolecules govern a myriad of cellular functions. Examples include cell growth, programmed cell death, sensing and metabolism. These events trigger complex cellular pathways characterized by reactions, environmental changes, intermolecular interactions and allosteric modifications. All of these processes involve molecular recognition, i.e. the process by which two or more cellular partners interact to form a specific complex. This recognition dominated by short-range, often transient, interactions at the contact surface of the molecules. Even conformational changes and assemblies of very large macromolecular aggregates, which can be propagated through long distances (few nm), are the effect of local interactions between small molecules (like messengers) or macromolecules with their cellular targets.

Ultimately, therefore, the integration of biological complexes into cellular pathways (the so called 'systems biology') requires a complete and quantitative description of these events (Mechanistic Systems Biology), so far mostly lacking. This impacts strongly on pharmaceutical sciences and toxicology, as drugs target (and mutations affect) pathways, rather than a single biomolecule. It is also crucial in nanobiotechnology, e.g. to design artificial sensing devices, which in nature involve cascade of events and not a single protein.

Structural prediction of a possible complex between Huntingtin N-Term fragment and F-actin

Expanding the scope of biomolecular simulation

Computational approaches are keys to face the challenges of mechanistic systems biology. Very broadly speaking, we distinguish here two major strategies. The first uses algorithms based on biological concepts such as Darwinian evolution, along with algorithms taken from the theory of information, the so-called structural bioinformatics: aligning DNA, RNA and protein sequences allows to predict 3D models of biomolecules and provide functional assessments of gene products. Among the many ways to do this, here we mention the comparative modeling approach based on the fact that proteins diverging from a common ancestor have similar 3D structures. Therefore, target proteins structure can be modeled using proteins with sizeable sequence identity as template. The approach is particularly crucial for structural predictions of membrane proteins, such as ion channels. Membrane proteins constitute 30% of the human genome and only 0.5% of the current protein structure database.
The second strategy is based on the laws of physics. Force-field based molecular dynamics (MD) predicts structural, dynamical and energetic (bio)molecular properties based on Newton's second law of motion. It is based on effective potentials. Several phenomena depend on the electronic states in such an intricate way that they cannot be modeled via effective potentials. Examples include enzymatic reactions, which involve bond-forming and bond-breaking phenomena, and optical properties such as fluorescence. More sophisticated and accurate approaches might be needed. In ab initio MD, interatomic forces are evaluated from electronic structure calculations (typically DFT) as the simulation proceeds. This finer level of description demands a much larger computational cost (0.01 ns, 100 atoms). As biological molecules are invariably large for a full first principle treatment, one may embed the quantum region in a classical region (the so called QM/MM method) following the idea of Warshel and Levitt back in the 70s. Most of these approaches require a high degree of parallel programming and huge computational resources to apply them to the large biological systems. Such resources are provided by both the most important supercomputing facility in Europe, located at Forschungszentrum Jülich, and the two local clusters managed by GRS.
As most biologically interesting processes and functionally relevant motions exceed by far present day capabilities of MD, statistical mechanics methods aimed at investigating rare events are widely used. Here, we are extensively using the recently developed metadynamics approach to evaluate kinetic and thermodynamic quantities of a variety of biochemical processes.

APPLICATION

Click here if you want to know how to apply for the Master's or Doctoral Program in Simulation Sciences at 
German Research School

For more information about the application contact us