Hi, I’m Charlotte Merzbacher. I'm a PhD candidate in the CDT for Biomedical Artificial Intelligence at the University of Edinburgh. My current research uses machine learning and differential equations to model genetically engineered bacterial cells.
I previously worked as a Data Scientist at Natera, a genetic testing company based near San Francisco. I graduated from Brown University with Honors in Biomedical Engineering.
I enjoy solving challenging problems at the intersection of biology, computer science, and mathematics. In my free time, I enjoy reading, writing, and hiking.
DATA SCIENCE SKILLS
Languages
Python
R, R/Shiny
Java
MatLAB
Manipulation
SQL
Bash Scripting
Pandas
NumPy
Analysis
Scikit-Learn
PyTorch
TensorFlow
OpenCV
Visualization
Matplotlib
Seaborn
ggPlot
Illustrator
EDUCATION
2021 to Present
Ph.D., Biomedical Artificial Intelligence
M.Res. with Distinction expected August 2022
University of Edinburgh
Machine Learning Practical & Theory, Mathematical Biology, Bioinformatics Algorithms
2015 to 2019
Sc.B., Biomedical Engineering (Honors)
Brown University
Sigma Xi Honors Society
Computational Biology, Reactor Kinetics, Statistics, Differential Equations, Algorithms, Biochemistry
April 2017 to May 2019
Research Student, Kane/Wong Lab
Voss Fellowship, Karen Romer Research Award
Profiled fibroblast phenotypes with image analysis
EXPERIENCE
October 2019 to September 2021
Production Engineer & Data Scientist
Natera
San Carlos, California
Managed operations on a team of 3 engineers for product algorithms with a total volume of >1M samples/year
Investigated lab issues using SQL and Python; completed over 500 investigations with an avg time to resolution less than 48 hours
Designed product metrics and alerts dashboards to predict and detect neural network and Bayesian algorithm issues
Developed new QCs to assess recoverable and irrecoverable lab failures, rescuing around 500 samples/month
Communicated analysis and algorithm results to interdisciplinary stakeholders (incl. biologists and genetic counsellors)
PROJECTS
Machine Learning Approaches for Optimization of Pharmaceutical-Producing Gene Circuits
Developing python package to design optimal genetic control architectures for metabolic engineering using ML
Single-Cell Detection in Brightfield Flow Cytometry Images
Applied supervised, unsupervised, and semisupervised methods to classify retinal cell images as singlets or doublets
Identification of BvgA-Dependent and Independent sRNAs in Bordetella pertussis
Performed a screen of the whooping cough (B. pertussis) genome for small RNAs related to virulence
Microbial Fuel Cells: Theory and Practice
Python-based models and experimental results from a yeast-based MFC. Report here
Integration of DNA Methylation Datasets for Individual Prediction
Identified normalization methods that project datasets from different populations onto each other with minimal bias
Honors Thesis: Profiling Fibroblast Differentiation in 2D and 3D
Designed python, pipeline to increase throughput and accuracy of biomarker quantification in 2D and 3D models of lung fibrosis. Improved image assessment speed by 15x while retaining similar accuracy rates to manual assessment
Jupyter-notebook based interactive textbook on evolutionary dynamics
R/Shiny app analyzing my reading habits from the past 5 years
NOTEBOOK
Useful Python for Data Science
In the two months since starting my job as a Production Engineer in the data science group at Natera, I’ve learned that having modular and reusable code is incredibly useful for when unexpected analyses and crises arise... read more
From BLAST to BEAST
Building phylogenetic trees from sequence data
The NCBI BLAST web service is awesome, but it can be difficult to set your exact algorithm parameters and parse the output to use in evolutionary analyses. In this tutorial, I’ll be taking you from a single sequence of... read more
Perfect Ginger Crinkles
I love baking, and I especially love ginger cookies. For years, I’ve been searching for a ginger cookie recipe that is easy, consistent, and tasty both fresh and a few days after baking. After a lot of experimentation... read more
built by charlotte merzbacher