IBM®
Skip to main content
    United States [change]    Terms of use
 
 
 
    Home    Products    Services & solutions    Support & downloads    My account    
IBM Research

Protein Folding

Biomolecular Simulation research at the IBM Almaden Research Center is focused on understanding the underlying mechanisms of protein folding. Our Group is an active member of IBM's Blue Gene project (http://www.research.ibm.com/bluegene), which is a research project to build a next-generation supercomputer and use it to study protein folding.

Proteins are linear polymers made from amino acid monomers put together in an order specified by the corresponding genes. A protein's amino acid sequence specifies its three-dimensional structure, which in turn defines its biological function. This includes its ability to interact with other molecules in the cell, its ability to carry out chemical reactions, and whether or not it can bind to specific drug molecules. Knowledge of a protein's three-dimensional structure, and how that structure can change, is important to fully understand a protein's biochemical role. Currently, we are focusing on the kinetics of folding to understand why some proteins fold up in microseconds while others take seconds or minutes to fold. By using Blue Gene, we can carry out many microsecond length kinetic simulations to study some of these fast folders.

At Almaden, we use Molecular Dynamics and Monte Carlo simulation techniques to study the protein folding process all the way from the completely extended or unfolded state to the final folded structure. We use detailed computer models of the protein and its environment that compute the positions and energies of each atom in the protein. We have successfully used these techniques to simulate the folding of small proteins (12-20 amino acids) from a completely unfolded state to within 1.0 Angstrom (*) of the experimental structure. We are currently applying these simulation techniques to study the folding behavior of larger proteins ranging from 40-100 amino acids, with a particular emphasis on proteins that show anomalous folding thermodynamics or kinetics.

Our research consists of four main areas:

  1. Developing and implementing algorithms for efficient simulation of protein dynamics.
  2. Simulating the kinetics and thermodynamics of folding for a range of increasingly complex protein systems.
  3. Analyzing simulation trajectories to extract kinetic and thermodynamic properties of the folding process.

Understanding the underlying mechanisms of protein folding will have major impact not just on biology and the life sciences but also on our ability to design similar nanostructured polymers.

Representative structures of the 80 amino acid lambda repressor 6-85 fragment at different temperatures from one of our thermodynamic simulations

Representative structures of the 80 amino acid lambda repressor 6-85 fragment at different temperatures from one of our thermodynamic simulations.





  

Images, click to enlarge
tryptophan cage
Simulation results for the 20 amino acid "trp-cage" peptide. The simulated structure is shown in green, superimposed on the experimental structure in red. The simulation starts from a completely extended structure and reproduces the experimental folded state to within 1.0 Angstrom (C-alpha RMSD)(*).
(*) C-alpha RMSD; RMSD = Root Mean Square Deviation.


    About IBMPrivacyContact