I've always been interested in lots of different pure scientific type questions that come up in biology, math, and physical science, which made it easy to decide to focus on science but hard to pick a specific discipline. Early in college, I figured out that I liked biology and math questions the most, and I eventually realized that you need computation to do research that combines the most interesting aspects of math and biology. Once I realized this, I was motivated to learn programming, which I hadn't been that interested in before.

• Devising summary statistics that will reveal new information from old datasets and ideally have useful mathematical descriptions
• Weaving a collection of results into a story
• Formulating explicit computational tasks that will that shed light on an ill-defined but important biological question
• Learning to anticipate the main complaints that various audiences will have about your work
• Judging how much perfectionism is needed for a given task

I had a really good undergrad education in pure math, learning a lot of algebraic geometry, complex analysis, etc. Although those topics don't really come up in my research nowadays, I think that doing proofs for those classes, as well as two pure math REUs, gave me the "mathematical maturity" that's integral to my research, i.e. stamina and strategies for working on tough theoretical problems. I didn't take many classes in statistics, but became fluent enough in reading math that I can usually look up whatever statistical techniques I need when I need them. I learned to program in python concurrently with starting research in computational biology, just using the internet to look up how to do the various computational tasks that I figured out I needed to do. It was a crude but effective approach; I get by fine, but I might have more efficient programming habits if I'd taken a class or two.

My research is motivated by the super-broad question of how biological forces like mutation, demography, recombination, and selection interact to generate the patterns we see in genomic datasets. I started off focusing on inferring demographic events like population size changes and migration, but switched gears to studying multinucleotide mutation events for awhile because they could explain some patterns that were bugging me in that I couldn't use demography to explain them. I hope to just keep studying the most interesting patterns I see in genomic data insofar as grant applications don't force me to do more planning ahead than that.