My work sits where molecular biology meets computation: wet-lab biochemistry on chromatin proteins, and machine learning for spatial gene expression.
UNITES Lab
Training models to predict spatially resolved gene expression (Visium HD, 8 µm bins) from routine H&E-stained colorectal cancer slides. I compare training objectives (mean squared error, flow matching, and zero-inflated negative binomial likelihoods) and study uncertainty, sparse counts, evaluation metrics, and generalization.
Strahl Lab · senior thesis
Investigating histone binding by Sas5, a subunit of the yeast SAS-I histone acetyltransferase complex, using protein purification, peptide-binding assays, and immunoblots. The binding mechanism is still an open question.
Collaboration with the Jacobsen Lab
TODO: one or two sentences on what was validated and how.