Welcome! Β‘Bienvenidos!

My portfolio and several tools are below. And pictures of my dog.

🧬Pre-publication genetic disease research

Pre-publication work on population prevalence and enhancing genetic diagnostics
Fullscreen

πŸŽ“Master's Thesis

It's my thesis. Objective: Characterize hemoprotein binding sites (e.g. hemoglobin, cytochrome) to inform protein engineering strategies. Data was extracted/generated with Python/UCSF Chimera, analyzed/visualized with R. Results suggest nonpolar amino acids in the binding pocket may have greater importance in heme binding than previously suspected.
Fullscreen

πŸ”¬πŸžStats, StreamlitπŸ”₯

Statistical analysis to identify prime candidates in a screening study. Spectrophotometry absorbance/time-series data were detrended, denoised, and statistically significant samples were identified. From thousands of samples of high-throughput data, approx. 100 candidates to move to next stage.
Fullscreen

Portfolio

This Website
This Website
August 2022 - June 2023
The page you're viewing and everything besides the Streamlit iframes above are implemented using Javascript/Typescript, React, with Nextjs and a heavily edited Mantine component library.
Technologies Used:
βš›οΈReact
TS/JS:TypeScript/JavaScript
β–²Node.js/Next.js, Vercel
πŸ”₯Streamlit
Master's Thesis
Master's Thesis
April – September 2021
Analyzing structural features of hemoprotein binding pockets to identify relevant features for heme cofactor binding.
Technologies Used:
🐍Python
πŸ΄β€β˜ οΈR
πŸ™€Github
UCSF Chimera
Preliminary Research, genetic disease
Preliminary Research, genetic disease
February 2023
Leverage genetic database, protein language model and present literature review in grant proposal.
Technologies Used:
🐍Python, Biopython
πŸ“ˆPlotly, Quarto/Jupyter
☁️Github CI/CD, Netlify
πŸ€–ESM1b Language Model, AlphaFold
Statistical Analysis/Candidate Selection
Statistical Analysis/Candidate Selection
Identifying prime yeast strain candidates from a pool of approx 6000 strains. Involved time-series analysis/detrending via linear regression, ANOVA to justify pooling samples across instruments, and p-value thresholding.
Technologies Used:
🐍Python
πŸ“ˆplotly/statmodels/scipy/matplotlib/seaborn
πŸͺGoogle Colab/Jupyter
πŸ”₯Streamlit (~Tableau)
SMILES Code to Target Class
SMILES Code to Target Class
July 2022; Updated Feb 2023
Predicting a ligand's target protein class. Machine learning models train on QSAR data generated from RDKit/Mordred.
Technologies Used:
🐍Python
πŸ€–ML: Sci-kit, XGBoost, PyTorch
πŸ™‚RDKit/Mordred
🐼Pandas
πŸͺGoogle Colab/Jupyter
☁️AWS EC2
RNA-Seq Analysis and Model
RNA-Seq Analysis and Model
May 2022
Predicting if a patient has breast cancer based off their RNA expression levels and clinical data.
Technologies Used:
πŸ΄β€β˜ οΈπŸ§¬R, edgeR
πŸ€–ML: Sci-kit

About my dog and me

Sherlock

He is the best. He has one white paw. A friend suggested naming him Michael Jackson.
  • Rescue boi – Sherlock is, according to this adoption papers, an Australian Kelpie/Mix.
  • The drama – I suspect he's part chihuahua, because he can be grouchy. Also he's smaller, and cuter than the dogs that come up in pure-bred photos.
  • Mexican? – He likes spicy food and he responds better in Spanish. I'm pretty sure he was a street dog in Tucson for a while.
  • Dog-shaped person – He enjoys long walks, car rides, and wearing fashionable clothing.

Pat

  • Software Engineering & Bioinformatics – My MS was in Bioinformatics. I've used a number of libraries and languages to achieve results: Python, JavaScript/TypeScript, R, for visualization, Quarto, Rmarkdown, libraries like Plotly, etc.
  • Pharma R&D, Wetlab – Before my Master's, I worked for 2 years in antibody formulation development. Having the science background has been invaluable in translating chemistry between empirical papers and the computational methods
  • Outside work – I like cooking, gaming, skateboarding and recently I've been playing with an Arduino
  • I too enjoy spicy food... like a delicious, purifying sauna: all the sweat and tears leaving my body as I nourish it.

Sherlock