Computational biology meets design. I build visualizations, simulators, and presentations for complex biological data. Reach out if you'd be interested in minimal, custom designed charts/graphs for a project!
Master's student in Applied Data Science
UMich · UC Berkeley · Seanome · UCSF · CZ Biohub
20 concatenated charts generated from Deidentified raw Excel data using Altair, visualizing age-associated changes in cutaneous biomarkers across vulvar disease states. Custom design built to support clinical researchers at Brown University, Providence, RI.
A stochastic model of B-cell antibody maturation built in pure Python. Implements a full IUPAC degenerate motif matching engine to identify AID hotspot sequences (RGYW, WRC, GNW), scales per position mutation rates by literature derived enrichment weights and immune cell state, and outputs annotated mutation logs with a hotspot pressure visualization.
CV review for 2nd interview at CZ Biohub. Covers heregulin stimulation experiments across HER3 KO cell lines to investigate whether HER2 self-phosphorylation drives downstream AKT and MAPK signaling. This challenges the idea of drugs targeting the extracellular domain.
Ethical case study analyzing MelanoBio, a hypothetical POC focused melanoma diagnostic AI, tracing bias accumulation from research literature through neural network training to clinical deployment. Recommends outcome based governance over pre-deployment filtering alone.
Journal club presentation reviewing Ni et al.'s dual fluorescence readthrough reporter system, which uses a genome wide CRISPR Cas9 screen across 18,053 genes to identify both known and novel cleavage and polyadenylation factors in human pre-mRNA processing.
Can simplifying the alphabet proteins are written in actually improve how well a model understands them? This journal club presentation for Seanome unpacks how reduced amino acid alphabets interact with protein language models like ESMFold
Interactive timeseries of synthetic biology part submissions across the SynBioHub.org registry. Queried via SPARQL, cleaned with pandas, rendered in Altair. Hover, click, drag.
Explore whether country level agricultural and livestock production patterns are associated with the occurrence of zoonotic disease outbreaks. By combining production statistics with case records, this analysis explores the data to uncover any patterns that may aid epidimeological characterizations of zoonotic outbreaks, before Covid 19 changed everything.