The projects below are representative examples of how I approach scientific software, reproducible research, and technical training. They are public projects that can be inspected directly rather than portfolio mock-ups.
tcpl — high-throughput screening infrastructuretcpl is an R package and data-processing framework developed for high-throughput chemical screening data at the U.S. EPA. I did most of the package’s design and implementation work while at the National Center for Computational Toxicology.
The project required more than statistical analysis: it involved database-backed data management, scalable processing, reproducible workflows, API design, user-facing functions, documentation, and maintaining a system that could be used by researchers beyond the original development team.
Relevant work: R package architecture, scientific pipelines, databases, high-throughput data, reproducibility, statistical computing.
toxpiR — scientific decision-analysis softwaretoxpiR implements the Toxicological Priority Index framework in R. I did most of the original software design and coding, translating a scientific methodology into a reusable analytical package rather than a one-off analysis script.
Relevant work: scientific software design, R package development, visualization, quantitative methods, reusable analytical tooling.
filer2020A — a scientific publication as a reproducible R packagefiler2020A demonstrates an approach I strongly favor for computational research: treating the complete analysis for a scientific publication as a versioned software project.
The repository is an R package containing analysis code, data, documentation, dependencies, and vignettes so the computational record of the study can be distributed and reproduced as a coherent unit.
Relevant work: reproducible research, package-based analysis, statistical workflows, documentation, publication-quality scientific computing.
I developed the How to Learn to Code / Practical R curriculum for graduate researchers at UNC and taught it across multiple cohorts, including remote instruction.
The course was designed around practical scientific programming rather than syntax alone: working with real data, developing reliable habits, debugging, structuring analyses, and helping researchers become independent R users.
Relevant work: R training, curriculum design, live technical instruction, documentation, mentoring.
My GitHub profile includes additional work in bioinformatics, copy-number analysis, visualization, statistical methods, utilities, and research software. My scientific publications are also available through Google Scholar and ORCID.
If you have a project that resembles any of these problems, see Services or email me.