Conduct and advance biochemical data analysis. Advance our computer aided drug design pipeline which comprises encoding novel molecular descriptors, implementation of specialized machine learning tools, statistical analysis of biological response data, and use of exploratory cheminformatics techniques. Contributed to pharmaceutical research through teams involved in bioinformatics and cheminformatics.Work in a team that develops and uses machine learning models to elucidate structure-activity trends and to predict which virtual molecules were most likely to exhibit desirable values for many properties. Review teammates' code and implement tests to ensure readability standards and robustness of function. Added functionality and explained practices based on findings in the cheminformatics, scientific, and machine learning literature. Data interpretation, data visualization, and data analysis in a biological, chemical, and pharmaceutical context. Develop a robust statistical testing platform for the detection of points in the tail regions of large datasets.Employ unsupervised learning to explore activity trends within isolated pockets of chemical space. Implemented an ML and statistical data mining workflow for accelerating chemical structure-activity dataanalysis to identify activity cliffs in a set of protein degraders.Shared findings with colleagues in meetings with interdisciplinary teams through oral presentations and written reports.

I conducted software development and data science for a bioinformatics research group. My former group conducts experiments as well as host web-pages that aid in processing and analyzing data from hundreds of thousands of researchers from around the world as a public service. Analyzed datasets derived from biological pathways. Visualized results and clusters of data from metabolomics datasets.My duties include writing software in C++, R, and Python using Rcpp, cmake, Make, git and github in the build process. I have also treated data-sets and developed models using scikit-learn and matplotlib.I am most interested in models of real-world systems and in software that relieves burdens on computational resources.

Programmed in C++, C, and Python using MPI and OpenMP to achieve parallelism. Submitted jobs to a high performance computing grid for research and homework using PBS scripts. Wrote a thesis on code that I wrote that reduced the number of calculations in Quantum Chemistry Kernels and eliminated the use of disk for workflows where equation of state (Fock) terms were larger than system memory. Used existing libraries in the codebase to accelerate Range-Separated DFT calculations by 50%. Reformulated DFT terms to reduce the memory footprint by 33%.Code was incorporated into the Psi4 quantum chemistry package and has been used by thousands of users worldwide.