PROJECTS — 3 TOTAL
Multi-Omics Integration Pipeline
SHIPPEDSnakemake pipeline connecting genomics, transcriptomics, and proteomics outputs. Built to handle the messy data we actually get, not clean benchmark data.
Protein Function Classifier
IN PROGRESSUsing ESM2 embeddings to annotate protein sequences from environmental samples. Still figuring out where it breaks.
Lab Notebook to Structured Data
CONCEPTTired of unstructured Benchling notes. Exploring whether an LLM can consistently pull structured records out of free-text observations. Early days.