Alan Grossfield grew up interested in math and science, with parents who worked in quantitative fields, as a math teacher and a structural engineer. “When I was about five or six years old, I heard my parents mention pi,” he shares. “To deflect my disappointment that we weren’t discussing dessert, my dad got a bunch of cans out of the cupboard, along with a piece of string and a ruler, and we measured pi. I remember the shock and glee I felt when the same number came up, over and over.”
As a high school student, Alan Grossfield started asking big questions in his advanced biology class. “We were studying how ribosomes worked, and I got really frustrated with the teacher. He was explaining that the tRNA did this, then the ribosome does that, and then the protein chain gets longer, and it kept sounding like the ribosome knew what it was trying to accomplish. I kept asking how it did it, since the molecules don’t have intent, and I got mad the teacher didn’t know,” he explains. “Of course, it wasn’t his fault — the first ribosome crystal structure wasn’t published until eight years later — but I came away convinced that I wanted to understand why molecules do stuff. I’ve never gotten past that — realizing that a bunch of molecules that only communicate by direct collisions manage to create function across nine orders of magnitude in length scale still blows my mind.”
He started at Cornell University in 1990 as a biology major. “I was already interested in biophysics, but I didn’t know the word,” he says. “I was very lucky that my assigned freshman advisor was Jerry Feigenson. Although that relationship is only official for freshmen, I went back to him for advice in the middle of my sophomore year. I was getting frustrated with my classes; I wasn’t learning what I wanted to know, and I was sick of all the memorizing I had to do in my bio classes. Jerry listened, and after a while he suggested I’d be happier switching to a physics major.” Grossfield did change majors at the end of his sophomore year and finished his degree in physics with a concentration in biology in 1994.
He then entered a graduate program at Johns Hopkins University which was then called the Intercampus Program in Molecular Biophysics. He was the first student in the lab of Thomas Woolf, using molecular dynamics simulations to study membrane-protein interactions, focusing mostly on the interactions between analogs of tryptophan side chains and the membrane-water interface.
In the late 1990s, Daniel Zuckerman, now a professor at Oregon Health & Science University, joined Tom Woolf’s group at Johns Hopkins as a postdoc late in Grossfield’s time there as a grad student. “Although nominally I was senior to Alan, in fact he was a mentor to me in terms of learning structural biology, membrane biophysics, and good computational practices,” Zuckerman shares. “Alan is very astute in his physical thinking and his ability to spot problems in data, which are really critical skills in our field.”
Following the completion of his PhD program in 2000, he started a postdoctoral fellowship at Washington University School of Medicine in St. Louis with Jay Ponder in the Department of Biochemistry and Biophysics. He had been struggling with computational limitations on the quality of statistical sampling, so he decided to change lanes to do modeling where statistical sampling was not a major problem. “I went to Jay’s lab to work on some protein structure prediction work involving potential energy smoothing, but once there I switched gears to study the solvation thermodynamics of simple ions using the AMOEBA polarizable force field, which was being developed at that time by Jay and another postdoc in the lab, Pengyu Ren,” he explains. “It was a very exciting time to be a computationalist at the med school. From Jay, I learned a huge amount about simulation methods, as well as how to design code to be clear and maintainable. The Center for Computational Biology had hired several outstanding new faculty members, including Rohit Pappu, Nathan Baker, and David Sept. I knew Rohit from my grad school days — he had a lot to do with my choosing to go work for Jay — but I found the volume and diversity of work going on in all of their labs inspiring, and watching them navigate the new professor experience informed my choices when I got my own lab years later.”
In 2004, he moved back to Yorktown, New York, where he grew up and began a second postdoc at the IBM TJ Watson Research Center. An informal mentor of Grossfield’s since graduate school, Scott Feller, had been collaborating there with Michael Pitman on simulations of lipid membranes, and recommended him for the position. “The team there had been developing a custom molecular dynamics code specifically designed to take advantage of the new Blue Gene supercomputing architecture. My timing was exceptionally lucky — I arrived shortly after the code was functional, and roughly six months before the Blue Gene/L supercomputer at Watson debuted at number 2 on the list of Top 500 Supercomputers (number 1 was another Blue Gene),” he says. “The projects Mike, Scott, and I worked on were allocated roughly onefourth of the machine, letting us run simulations far longer than was generally possible at that time. We used this opportunity to study the mammalian dim-light receptor rhodopsin, looking at its lipid-protein interactions, the role of polyunsaturated fatty acids in controlling function, and the early stages of activation. We also worked on better methods to quantify the quality of statistical sampling.”
Grossfield was offered a faculty position at the University of Rochester Medical Center while at IBM Research, where he now works. He is an associate professor in the Department of Biochemistry and Biophysics. “We have several established project areas, including simulating the early events in rhodopsin activation and comparing them to time-resolved X-ray scattering experiments, as well as developing methods to compute and interpret terahertz spectroscopy experiments on protein crystals,” he shares. “We’re also moving into new areas, including understanding the mechanisms of new potential drugs to treat opioid overdoses, using the experimental fibril structures of alpha-synuclein to think about the origins of Parkinson’s disease, and developing a new framework to think about the thermodynamics of phase separation in lipid membranes. We also put a lot of effort into developing LOOS, our suite for the analysis of molecular dynamics simulations.”
The most challenging period of Grossfield’s career has been over the last four years, during which time the National Institutes of Health (NIH) grant he had was not renewed and new applications to NIH and the National Science Foundation were not funded. He asked his department to pay his graduate students and postdoc, and was unable to take on new students, let alone upgrade lab equipment or pay to attend conferences. He petitioned his dean to extend his tenure clock, eventually obtaining a three-year extension. “Happily, things have finally turned around over the last few months. During the past year, I wrote nearly a dozen grant proposals, several on new topics for my group. Three of them were funded — two with collaborators, one just for my group, all in relatively new research areas for me — which is positioning me to rebuild my group and apply for tenure,” he shares.
James Seckler, Case Western Reserve University, was one of Grossfield’s unofficial mentees, and the two continue to collaborate. “At present, we are working together on developing a drug to reverse opioid induced respiratory depression (the cause of death in heroin and fentanyl overdoses) without affecting the ability of opioids to stop pain,” Seckler says. The most valuable thing he has learned from Grossfield over the years comes in the form of a saying he frequently shares in the lab: “There is always a faster way to get the wrong answer.” It’s his way of reminding everyone around him to slow down and be careful. He always stressed choosing techniques based on reliability rather than the time investment. This is something I took to heart working with him and continue to use to this day.”
Grossfield believes that his most important contributions to biophysics will be the students and postdocs he helps train. He offers this advice for young biophysicists as they build thieir own careers: “Invest early on to build a network of good mentors. When you’re picking a lab for your thesis or postdoc, pay attention to what kind of mentor the PI is likely to be. Seek out other mentors as well — there doesn’t have to be a formal relationship, just a recognition that this is a person you can go to for advice, and who is invested in your success. Some of my most influential mentors are people I never worked for,” he says. “On the flip side, look for opportunities to provide mentoring for others. You don’t have to be very far along your career path to do so; a second- or third-year graduate student can be an influential mentor to an undergrad or first-year graduate student. However, before you give advice, be sure you understand what they want. Your role as a mentor isn’t to push them along your path, but to help them identify their own goals, and then reach them.”