promor
An R package for proteomics data analysis and machine learning based predictive modeling
This project was perhaps my most ambitious one to date. When I first started working as a Computational Biologist at the Eastern Virginia Medical School, I had just made the transition from doing work that involved both wet lab and computational work to purely computational biology/ bioinformatics. I, along with two other colleagues, were hired to work primarily on a project called the Digital Patient at EVMS. Within the overarching goals of that project, we had the opportunity to design our own projects that would be of interest to the faculty at the school and would benefit their ongoing research. While working with a research group exploring proteomics-based biomarkers for early detection of prostate cancer, I thought of developing a custom bioinformatics pipeline to cater the group’s biomarker research. promor was built based on those initial efforts. With no experience whatsoever in software development of any kind, let alone R package development, I had a lot to learn about the best practices in the process. Looking back, I am surprised at how quickly the package came together - within just a few short months, promor made it to CRAN and a publication was in the works. Now, I am excited to see the package being used by research groups from around the world, and recently, promor passed 7000 downloads on CRAN.