Nathaniel Charest Email & Phone Number

Durham, Nc, Us

This role involves research into improving methods and developing tools relevant to the Agency's need to predict properties from structures. The work involves leading teams in the development of software to aid the efforts of analytical methods such as non-targeted analysis in mass spectrometry, facilitate chemist's ability to visualize data, perform feature engineering to improve toxicity and property models, and mathematically characterize chemical spaces relevant to the Agency's analytical interests.

Durham, North Carolina, United States

I subcontracted through Oak Ridge Associated Universities (ORAU) to assist the EPA in its research.My position involves assisting the development of state-of-the-art quantitative structure-activity relationship (QSA/PR) models for predicting chemical properties across multiple endpoints. These models are deployed on various computational platforms, the construction of which requires expertise in multiple software and scientific domains as well as collaboration across multiple fields.This resulted in a publication co-first authored with Charles Lowe, which was published in Chemical Research in Toxicology. https://pubs.acs.org/doi/10.1021/acs.chemrestox.2c00379

The EasyText.AI project consists of a team of interdisciplinary, global experts working to improve the accessibility of COVID-19 text information for the intellectual disadvantaged. Much of the information disseminated by government and private sector entities around COVID-19 is not readily accessible for the intellectually disabled. The goal of this project is to develop an artificial intelligence model for taking in text information and simplifying it into a more accessible form.As part of Phase 1, we are developing a crowdsource application for collecting a training set of data for our models. My work has involved architecting the integration of APIs for automated collection, cleaning and flow of text data, as well as developing the front end of a PyBossa framework application. I am managing the development of the crowdsourcing platform. We are hoping to launch within the United States quite soon! We are also pursuing better means of quantifying text accessibility and applying state-of-the-art NLP models for automatically simplifying text data.Please visit the team's website for more information.https://easycovid19.org/project/

Santa Barbara, California, United States

Continuing & finishing projects started during my work with the Shea group, my postdoctoral position is focused on finishing my work in using information theory approaches to predict the functionality of RNA molecules. Broadly, the theory being developed uses a first-principles mathematical model of epistatic interactions to predict regions of high activity within the fitness landscape. These techniques would have direct application in functional polymer engineering. I am currently working with experimental collaborators to explore the degree to which we can predict activity outside the training space of the model.I am also writing a review of machine learning across the domains of chemistry.

I pursued a Ph.D. in Computational Chemistry to satisfy my interest in cross disciplinary approaches to biochemical problems. My field of computational biophysical chemistry allows me to incorporate an array of approaches to my work and promotes deep, spanning knowledge of broadly applicable computation and theoretical skills while remaining grounded in the pragmatic realm of biophysical problems.My computational research on biopolymer systems was under the advising of Dr. Joan-Emma Shea. I used and tuned molecular dynamics models, performed data analysis using deep learning and traditional methods, while researching and applying novel machine learning techniques to biochemical computation problems. Developed skills in Python programming language, keras deep learning framework, data processing, and information theoretic approaches to data analysis. Additionally developed skills in molecular dynamics simulation theory, statistical mechanics and associate mathematical machinery and underlying physics.Resulted in the publication of "The Classifying Autoencoder: Gaining Insight into Amyloid Assembly of Peptides and Proteins." Another paper is currently in submission, under review.Current work involves the quantification of epistasis in ribozyme fitness landscapes using information theory. This will yield a fourth paper.

Santa Barbara, California, United States

I have consistently TA'd for classes as part of my graduate work. I enjoy teaching, particularly the challenge of conveying complex, abstract material whilst requiring that I convey it clearly to audiences with varying backgrounds and affinities. Leading discussion sections around more advanced topics improved my lingual and theoretical mastery of hard science and dramatically improved my professional communication skills. My responsibility with directing laboratories sections refined my managerial style and ability to guide project development, particularly when project design was part of the curriculum.General Chemistry Laboratory (Lower Division, Undergraduate Level)Biophysical Chemistry Laboratory (Upper Division, Undergraduate Level)Thermodynamics Theoretical (Upper Division, Undergraduate Level)Statistical Mechanics Theoretical (Upper Division/Graduate Level)Molecular Dynamics Theoretical & Lab (Upper Division/Graduate Level)Duties included managing multiple small projects by students, the creation of pedagogical materials, conveyance and enforcement of relevant safe practices, substituting in as a lecturer when necessary, as well as grading and proctoring.

Santa Barbara, California, United States

This was my first step into the field of computational biophysics and defined my interest and passion for both the field and its theoretical machinery. My work in this role informed my decision to pursue a Ph.D. in Chemistry.Performed molecular dynamics simulations, model validation and non-parametric statistical analysis for Dr. Joan Shea's computational chemistry group. Developed skills in Linux and Bash programming, as well as WHAM algorithm to approximate non-parametric probability distributions and perform analysis thereof. Developed knowledge of LAMMPS simulation software and operational knowledge of C++ (the source code of LAMMPS).Resulted in publication of "Systematic derivation of implicit solvent models for the study of polymer collapse"

Ph.D.

GPA 3.9 My thesis is in Model Development and Validation for Systems of Biochemical Polymers. Please contact me for a full CV.

Bachelor Of Science (Bs), Biochemistry

Activities and Societies: UC Regent ScholarGPA 3.6