Markov State Models for Molecular Dynamics
Using probabilistic graph models for efficient simulation of the protein folding process
Computational Protein Folding
Proteins are the machinery of life. They are comprised of a long string of amino acids, whose physical interactions wrap together into complex shapes essential to determining the protein's properties. The protein folding problem is about figuring out the 3-D structure of a protein given its 1-D amino acid sequence.
I reported how the Gridcoin cryptocurrency could incentivize people to donate computing time to scientists working on problems like protein folding. It was incredible that I could be helping cure cancer from my dorm room just by running a program on my laptop. Protein folding first instilled my interest in high performance computing and scientific research. I still remember purchasing an NVIDIA 1080ti graphics card for $1100 in 2017 because I could use it for protein folding, deep learning, and a little bit of gaming...
As of 2021, I have over 25 years of computing time donated to Rosetta@Home (University of Washington), WorldCommunityGrid (IBM), GPUGRID (Universitat Pompeu Fabra), and Folding@Home (Stanford). Even under optimistic assumptions, purchasing that compute from Amazon Web Services would net you a bill running into the thousands of dollars.
Data science and machine learning have made huge gains in computational protein folding research over the years, and contributing to this field has always been a (pipe) dream of mine. So far, though, I've only been able to write about it at school:
Komen for the Cure: A new campaign against cancer
This was a paper I wrote as a freshman undergraduate which ended up becoming the leading article in Stern’s annual “Call for Corporate Action” publication (mirror here). My research for the paper is where I first got interested in the quantitative sciences - before that, I was fully intent on pursuing a path of “business” and “entrepreneurship” (whatever that meant to my addled freshman mind). It’s a quaint and nostalgic experience to read my past opinions, so I like keeping this around as a reminder of where I came from.