IT News & Events

News about IT at Indiana University and the world

Menu

On the road to a cure for HIV using Big Red II and Karst

IU graduate student, Hao Sha in the Department of Physics at IUPUI is conducting exciting new research aimed at increasing the understanding of the biological mechanisms in terms of the fundamental laws in the physics and chemistry.

HIV Simulation Results

An example of output from Big Red II, Karst simulations

IU graduate student, Hao Sha in the Department of Physics at IUPUI is conducting exciting new research aimed at increasing the understanding of the biological mechanisms in terms of the fundamental laws in the physics and chemistry. “My current research aims to probe the thermodynamics of the assembly process of HIV capsid through coarse-grained simulations. I hope this research will offer useful information to help eventually find a cure to AIDS,” said Sha. He further explained, “HIV capsid is a proteinaceous shell surrounding the viral genome. It plays an essential role in the viral lifecycle. In recent years, the capsid has been established as a promising inhibition target, given that ill-formed capsids usually diminish or abolish the viral infectivity. However, the mechanism that governs the capsid assembly process is not entirely understood. In our work, we adopt a computational approach, to explore the thermodynamics of the capsid assembly. In particular, we parameterize the Lennard-Jones type potential for certain critical residues and reproduce the desired binding affinity between a dimer of C-terminal domains which are pivotal in forming the capsid. The methodology developed here along with the parameters determined lay a foundation for us to eventually simulate the entire capsid including more than a thousand of proteins.”

Hao Sha

Hao Sha, IU graduate student

In order to do his research, Sha has been utilizing IU’s supercomputing resources. He said, “Big Red II and Karst have been our primary platforms for conducting simulations. Without the high-speed supercomputers like BRII and Karst, it wouldn’t be possible for us to model the complicated biological systems such as the HIV capsid in accessible time. In particular, the simulations reported in our recent paper were all performed on BRII/Karst using CPU nodes. Currently, we are modifying our program to allow our simulations run on GPU nodes for better performance.”

 HIV Simulation Results

Another example of output from Big Red II, Karst simulations

Sha added, “The major challenge in my research area is how to obtain sufficient samplings in accessible time. Simulating large biological systems are time consuming. Proteins like the HIV capsid protein are consist of thousands of atoms, and the assemblies may further include hundreds of such proteins. Microsecond time scale is almost unreachable using all-atom simulations. Utilizing high speed computer clusters in combined with advanced sampling and modeling techniques can help to alleviate the challenge. By using coarse-grained model and running simulations on CPUs on BR2/Karst, we have already made it possible to obtain sufficient statistics for a dimer of HIV capsid proteins in ~14 days. Our ongoing research employing GPUs may further speed up the simulations.”