09.17.10 - Permalink
"Having access to NCSA kept my projects alive!"
By Barbara Jewett
Although variations of the gratefulness theme are frequently expressed when researchers share information about their work conducted using NCSA resources, Mihaela Bojin's appreciation goes a little deeper than most. Without access to free supercomputing clusters, she most likely would have abandoned her amino acid research for several years.
Like many young researchers, a few years ago Bojin faced an important decision: how to balance a research career's demands with those of a young family. Her post-doctoral research position was coming to a close and she had a toddler. She wasn't sure the rigors of academic research to build a career at a major university would afford her the amount of time she wanted to spend with her child, even with the assistance of a supportive and helpful husband. She opted instead for an assistant professor position in the chemistry department at Queensborough Community College (QCC) in Bayside, New York, part of the City University of New York system.
Queensborough in generaland the chemistry department in particularhas a research focus not typically found at community colleges. The chemistry department even boasts a new laboratory devoted solely to faculty/student research projects. But as is the case with so many colleges and universities, QCC lacks research compute power. For a computational scientist like Bojin, this is an obstacle. Enter NCSA.
Bojin joined Queensborough in 2007, knowing she wanted to expose her students to computational science. In fact, she started training students her very first semester as a professor, even though she had no funds.
"I had them read articles and we would discuss what is computational chemistry in very general terms. And luckily it crossed my mind," says Bojin, "that I used to have a grant from NCSA when I was a post-doc a few years before I joined here, and I remembered that it was fairly easy to get a grant. I thought, 'I need to apply for that to get my project started.' So I applied for a grant and I got it very fast. And then we actually got to run calculations and try our ideas, to try them out on the clusters. So for me, the program has been super essential. I don't know how I can be more grateful for it. It has literally been my start-up funding."
Bojin and her undergraduate team of two students are using NCSA's Abe and Cobalt clusters. Her teams have also run on the now-retired Tungsten. They are exploring hydrogen bonding in amino acids, specifically what happens to those bonds when the acidity or basicity of the environment changes. Think of a protein as being a Lego, she says. Amino acids are just the little building blocks in the Lego.
"I am looking at each of these small pieces to see how many forms they can take, because each of them is actually very flexible," she explains. "So they can move and they can form distinct conformations."
When you change the acidity of the medium of the amino acid, how does its internal bonding change? How influenced are they by the medium? These are questions Bojin hopes to answer.
"Amino acids are part of proteins, and proteins are part of our body. So it is very important to know how proteins work to figure out how healthy we are," says Bojin. "We are thinking that if we go to the smallest part of the protein, these amino acids, then maybe we can expand our knowledge to the bigger proteins. How proteins are changed when the acidity of the medium changes. It has been proved that slight changes in the medium in the protein can lead to mutations, and mutations can lead to, for example, cancers and all kinds of diseases. So it is important to know about all of these changes, then you can learn maybe how to prevent them or how to treat something like this.
"That is very far from our actual research. Our research is more on the fundamental level. The ultimate long-term goal is to correlate these changes in acidity for the small amino acids with what happens in an actual protein."
Amino acids' intra- and intermolecular interactions are significantly influenced by surrounding residues and the pH of their environments. Hydrogen bonds, in particular, control folding in secondary and tertiary structures in proteins and significantly affect enzymatic activity, Bojin explains. By employing density functional methods her team calculated different conformations of asparagine, arginine, glutamic acid, serine, and threonine in their neutral (no charges), zwitterionic, acidic (protonated), and basic (deprotonated) forms, in gaseous and aqueous media. They found that changes in acidity critically influenced and limited hydrogen bonding patterns, and thus the stability of the resulting conformers. The team is currently working on glutamine, aspartic acid, histidine, and dipeptidic conformers.
Bojin presented her research at the fall national meeting of the American Chemical Society (ACS) in both 2008 and 2009. Four undergraduate students also gave presentations at each meeting, and two more will participate the 240th ACS meeting in Boston in August. Bojin's students also presented their work at the ACS Mid-Atlantic Regional Meeting in 2008 and 2010, and at the Collegiate Science & Technology Entry Program (CSTEP) meeting in 2009 and 2010. At the latter her student Daniel Sangobanwo won first prize in the poster competition. In addition, her students give presentations every year at honors conferences and undergraduate research symposia of the ACS-New York section.
At a time when the science community is stressed to include computational science in a traditional four-year undergraduate college science education, where it is so difficult just to get students through the basics, Bojin's introduction of computational science to her community college students is a bit remarkable. And although it requires a significant time investment on her part, it also meets her need to keep active in research. The computational science she teaches complements what she does for own research.
"Before I came to Queensborough I was doing only research, and it would have been very bad if I had been disconnected completely from research work," she says.
Bojin says overall the students she teaches at QCC "lack a lot of the background, you know, not just for physical chemistry but background in chemistry and math and physics." She tries to give them the main idea about computational chemistry. And those who are interested she works with one on one.
"We don't have enough students whose interest is that sustained to try and do computational modeling with a larger, class-sized group. In fall 2008 I had regular meetings with four students who were interested, but at the end of the semester I ended up with only two of them who wanted to do research. It was kind of disappointing."
She says she has since found that only two research students at a time is better as she has a heavy teaching load. Once her students have a grasp of the method she then begins having them use the NCSA computers.
"I try to encourage them by giving students projects that are not very advanced in terms of the chemistry that they are using, pretty much molecule projects, things that they can quickly understand. We usually just run things in Gaussian so it is not very fancy," she says. "Then we have little workshops on using the program, and they start on their own making files and running them. They look together at the output. It's a process, so it takes a little while until they get things done," she says.
Bojin says her all her computational students have been motivated by the desire to move on to a good four-year school. They knew that doing research was something that was not only interesting but would give them an edge toward that goal.
"And they love computational chemistry because they have more freedom than they have if they work in the lab," she laughs. "In the lab you have more time constraints whereas if you work on the computer you can be on your own and still do work and finish up your project on Sunday night if you want. You can still do something if you are not here. If they have to travel in the winter break or the summer break, at least they can still do something because they have their laptops."
The Queensborough chemistry department is active in terms of research, notes Bojin. In addition to faculty and student research projects, faculty take students every year to undergraduate research meetings, to regional ACS meetings, and once a year to an ACS national meeting. It is easy for students to have a poster or an oral presentation for that kind of motivator, she says. And it gives them a deadline for their project.
Bojin says when working with undergraduate students work progresses more slowly, and there is not as much progress as needed to have a publication.
"We've studied five amino acids as thoroughly as we could and now, if we put them together, we could make a case. And hopefully a case that works. The work still needs to be finished up. I hope this summer I can get a manuscript draft completed. We are not the first ones to come up with the idea of studying these things. What we are probably the first ones to do, we are the first ones to systematically study every system, to study every case. Even in previous articles, researchers have studied little bits of each of these systems, but they haven't been put all together. If we look at these five cases we have a little story we can tell about the variations in rigidity of the systems and how they change when you immerse them in solvent. It comes out as a more round story than it was even half a year ago."
While work with these amino acids is nearing completion, she is working on two more projects with other students. None of it would be possible, she says, without NCSA.
"Given the limited time I have, I don't get a chance to run many calculations so the startup allocation limits are perfect for me. My runs are very little compared to a proper research group where you have things going every day. In my case, my students have something once or twice a week."
But that is more than enough time to inspire them with computational science research.