research mentorship

How I approach mentoring students in computational research.

Research mentorship plan

I work at the intersection of statistical computing, machine learning, and biological data analysis. My research has involved testing network inference methods, working with microbiome count data, and using high-performance computing to run large experiments. These projects have taught me how much research depends on decisions that are easy to overlook: what data to use, which comparison is fair, and whether someone else can reproduce the result.

I want students I mentor to become comfortable making and explaining those decisions. A successful project might produce a paper, a software package, or a useful analysis. It should also leave the student more able to frame a question and carry it forward independently.

How we would work

We would start with a question small enough to test. Before building a new method, we would read relevant work and establish a baseline. If an experiment gives an unexpected result, we would inspect the data, code, and assumptions before deciding what it means. Sometimes the result will be that an idea does not work. That is still worth documenting clearly.

I would ask students to keep a short research log with links to code, figures, decisions, problems, and next steps. We would use it to make meetings concrete. The meeting schedule and level of direction should fit the student and the project; I would not expect a new researcher to work with the same independence as someone preparing a thesis.

For computational projects, we would keep code and instructions in version control, record how results were produced, and review changes together. Students should be able to ask about a statistical assumption, a programming choice, or a confusing paper without feeling that they ought to know the answer already. I would give direct feedback on the work and make time to explain my feedback.

What students can expect from me

I would help define a tractable project, point students toward useful reading and tools, and review their code, analyses, figures, and writing. We would discuss authorship and credit early, then revisit those conversations as the work changes. I would support students in presenting their work and in choosing the next step that fits their goals, whether that is further research, graduate study, or another technical path.

I cannot promise a publication, funding, or a particular outcome. Those depend on the project and the institution. I can promise to be candid about progress, expectations, and what support I am able to provide.

What I ask of students

Bring your questions and unfinished work to meetings. Tell me when a task is taking longer than expected or when an explanation has not helped. Keep enough notes that we can retrace an experiment, and give collaborators credit for their contributions. As a project develops, I will ask you to make more of its decisions yourself and explain the evidence behind them.

If you are interested in working together, email me with the question you would like to explore and a little about your experience. A clear question is more useful than a polished proposal.