Experience

The Jackson Laboratory

• United States
• Biotechnology Research
• 700 & Above Employee

• Data Scientist

  • Mar 2023 - Present

  Developing methods for computational metabolomics and automated metabolome assembly. Developing methods for computational metabolomics and automated metabolome assembly.



Los Alamos National Laboratory

• United States
• Research Services
• 700 & Above Employee

• Postdoctoral Researcher

  • Feb 2021 - Sep 2022

  Postdoctoral research associate in the CCS-3 (Information Science) and C-CDE (Chemical Diagnostics and Engineering) departments. My main responsibility was developing C-CDE's metabolomics capabilities and to apply the resulting pipeline to a large multi-disciplinary project that was attempting to improve the drought tolerance of Maize under through the directed evolution of the soil microbiome. I was responsible for experimental design, sample collection and preparation, analysis using two-dimensional gas chromatography time-of-flight mass spectrometry (LECO Pegasus GC-HRT+ 4D) and/or enzyme assays, data analysis, and interpretation. This work identified several metabolites and implicated a key metabolic pathway we hypothesize contributes to the enhanced drought tolerance induced by the selected microbiomes. This work was awarded a LANL 2022 Spot Award for "initiative or creativity in addressing a critical need or difficult problem". Additionally, I worked on several side projects. The main side project was using machine learning to build models that can predict the presence of chemical substructures from electron ionization mass spectrometry (EI-MS) spectra. I then used the predictions from these models to improve assignment accuracy in a robust, measurable manner. This required building individual supervised models for each substructure and then using genetic algorithms to determine the ideal combination of models for improving assignment accuracy. Other side projects included a robust non-parametric statistical analysis of a dataset examining iodide uptake in various soils, the use of SPME to study volatile plant compounds, an analysis of the impact of soil wicks on metabolite and nutrient distributions in pots, and a metabolomic analysis of plants grown in the absence of a soil microbiome. Finalist in LANL's "Science in 3" scientific communication competition. Multiple publications in development, one currently in review.



University of Kentucky

• United States
• Higher Education
• 700 & Above Employee

• Medical Student Researcher

  • May 2019 - May 2021

  Using the tools I developed during my PhD, namely SMIRFE and the lipid classification model, I identified lipid profiles associated with the development of non-small cell lung carcinoma. I also further developed the graph theory-based tools for chemometrics to aid in metabolite database harmonization. This work led to a first author publication and a second-author publication.

• Graduate Student Researcher

  • May 2014 - May 2019

  I developed several novel algorithms to solve outstanding problems in the fields of metabolomics and analytical biochemistry. One was the development of an efficient algorithm for the untargeted assignment of molecular formulas to spectral features observed in Fourier transform mass spectrometry datasets. Another was a random forest model to predict lipid categories based on assigned molecular formulas from the previous tool. Additionally, I developed a chi-squared derived metric to identify and remove spectral artifacts in mass spectrometry datasets that confounded downstream machine learning models. This work led to several first-author publications and a US patent. Additionally, I also aided in system administration tasks in the lab such as the virtualization of infrastructure, maintaining computer systems, and the selection of new hardware. I was the go-to lab member for code and computer performance optimization.During this time I became familiar with a number of different technologies including: Python (including Cython, Sklearn, Scipy, Numpy, and multiprocessing), Perl, R, C++, Object-Oriented Design, SQL, Algorithm Design, Software Testing, Performance Optimization, Unix/Linux, Virtualization (VMWare, Qemu/KVM), Git, Docker, Puppet, Computer Hardware and Networking.



University of Louisville School of Medicine

• United States
• Higher Education
• 300 - 400 Employee

• Medical Student Researcher

  • May 2012 - May 2014

  I developed an efficient algorithm that solves the maximum common sugraph isomorphism problem in the context of chemical structures thus enabling fast chemical substructure searching. This work was published as my first first-author publication. During this time I became very familiar with Perl, mutiprocessing and Linux. I developed an efficient algorithm that solves the maximum common sugraph isomorphism problem in the context of chemical structures thus enabling fast chemical substructure searching. This work was published as my first first-author publication. During this time I became very familiar with Perl, mutiprocessing and Linux.



University of Louisville

• United States
• Higher Education
• 700 & Above Employee

• Undergraduate Research Assistant

  • Dec 2009 - May 2012

  During my undergraduate time at the lab my primary task was learning how to code and becoming familiar with algorithm design, object-oriented design and Linux. I familiarized myself with complex data structures and became very knowledgeable on graph theory and representations of chemical structures as graphs. I began work on an graph-theory based algorithm for chemical structure comparisons. I was also the leader for our university's entry in the Mathematical Competition in Modelling. Our team won "Outstanding Winner" (highest placement) in the Mathematical Competition in Modeling (MCM) for Discrete Mathematics (2012). This is an international coding competition involving over 3500 teams. Our solution required using the laboratory's HPC resources.