This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
With this award from the Major Research and Instrumentation (MRI) program, Adam Van Wynsberghe, Wei-Jen Chang, Natalia V. Connolly and Nathan Goodale of Hamilton College will acquire and operate a high-performance Beowulf-style computing cluster. This computing cluster will enable faculty in the departments of chemistry, physics, biology and anthropology to employ cutting edge computational technology in their research labs and classrooms. Research projects to be investigated include studies of the eukaryote tree of life, development of new tools for the study of dark energy, studies of infrared galaxies, examination of ribonuclease catalysis and a survey of influenza neuraminidase inhibitors.
A computing cluster is a group of linked computers that work in concert to achieve vastly more computational power than the individual machines. These clusters make it possible to solve complex problems in areas such as molecular dynamics simulations and statistical analysis that require extensive computational power. The computational infrastructure made available by this award will be used in sophisticated research projects and in courses across a variety of disciplines introducing numerous undergraduate students at Hamilton College to scientific computing.
Under this NSF award, Hamilton College has established and demonstrated a model for a college-wide, shared-use high performance computing cluster at a primarily undergraduate institution. Given the cost of purchasing and maintaining high performance computing, the policies and infrastructure we have developed to make this resource readily available to both major and minor users are a useful step in efficiently utilizing public and private funding. The public investment by NSF in this award has spurred approximately $100,000 in private investment from Hamilton College in the computing resource as its effectiveness as a campus-wide facility has been recognized. In addition to the PI and co-PI's on the grant, four additional faculty led-groups from Hamilton, in the Economics, Psychology, Mathematics, and Geology departments, have used or will soon be using the resource. Most importantly at a liberal arts, undergraduate-only institution like Hamilton, 33 undergraduates and 4 post-baccalaureate students have been trained by and worked on the projects enabled by this NSF award, and over 100 students have learned from and gained experience in the techniques used in these projects in their undergraduate classes. Seven students from the PI and co-PI's labs are pursuing graduate degrees in science (5 Ph.D.; 2 Masters) and two students have been awarded prestigiuous NSF graduate fellowships. The primary scientific work has taken place in the biology, archaeology, and chemistry, disciplines: Biology We were able to systematically test whether a class of molecules, the ribosomal proteins which are universally present in virtually all life forms, is suitable to help infer evolutionary history of eukaryotic organisms. Our results showed that ribosomal protein sequences suggest a phylogeny similar to those proposed by other researchers using a larger amount of sequencing data – a potential value of reducing needs of larger datasets and computing time. Our results also indicate that ribosomal proteins in species without mitochondria, the powerhouse of cells, have evolved faster than those in species with mitochondria. This phenomenon was first predicted approximately 23 years ago and we just had the ability to show that the prediction was correct. We also used this "seed cluster" to convince Research Corporation that we have the ability to form interdisciplinary research among faculty and analyze large datasets. Research Corporation funded two Co-PIs on this grant with an award to investigate gene expression networks in the parasite that causes the white spot disease in fresh water fish in 2012. We also benefited from funds from Hamilton College in recognition of this grant that enabled us to invite leading researchers in genomics and bioinformatics to give talks at Hamilton. Finally, the cluster also provided us with a unified platform to help teach a course in Bioinformatics more effectively. Archaeology The Goodale group has been researching stone tool technology used in historic and prehistoric contexts in Ireland, the Middle East and the interior Pacific Northwest of North America. We are using a variety of methods for analysis from portable X-ray fluorescence to measure the elemental composition of stone and mathematical models of morphology to examine potential evolutionary relatedness. With both types of analyses we are able to gain valuable insight into how humans used stone technology in the past. Archaeology majors at Hamilton College have greatly benefited from the HPC facilities. Four of the six students that have been part of the Goodale group are either attending or have been accepted to graduate school and one other has a teaching fellowship for the 2013-2014 academic year. Chemistry Our work has focused on understanding how small molecules, i.e. drugs, interact and bind with their protein targets. Specifically, we have investigated how small molecules bind to neuraminidase, a protein essential to the influenza replication cycle. We have created a multi-scale method that allows us to visualize the entire binding pathway, from observing the diffusive motion when the protein and ligand are distal to describing the multitude of short-range interactions that occur when the protein and ligand form their final bound complex. Our results suggest that the diffusive trajectories of the ligands are strongly funneled toward the active site, but following this initial approach, there are many various pathways and ordering of events in which the ligand makes its final bound contacts. These observations serve to deepen our understanding of small molecule-protein interactions and have the potential to inform the rational design of new influeneza antivirals. In addition to this work, we have undertaken projects that explore how protein mutations affect protein-protein binding and enzyme catalysis. We have also developed educational exercises for three undergraduate laboratory courses. Using our experience with protein visualization, we developed a biochemistry laboratory experiment utilizing the software VMD; we have used the cluster to implement a quantum chemistry investigation in our physical chemistry laboratory; and we have developed the tools to enable a protein-ligand docking exercise for our freshman chemistry course.
