Transcriptomic Cell-type Heterogeneity

Linking gene expression to functional DNA repair, cell-type context, and prognosis.

Brief

Transcriptomic profiling complements functional assays of genome integrity by revealing gene-expression programs tied to DNA repair capacity and cellular context. Highlighted here my ongoing research and completed coursework projects connecting expression data to repair phenotypes, cell-type composition, and prognosis.

Projects

  • PBMC expression ↔ functional DNA repair (DRC)
    Relates bulk PBMC gene expression to FM-HCR–derived DNA repair measures, with contrasts across PBMC subtypes to probe cell-context effects on repair phenotypes. Also profiles DNA repair pathway genes across blood cell subtypes to contextualize functional differences observed in repair assays.
    Status: Manuscript in preparation (contributing/second author).

  • NER gene expression in blood cells
    Compares nucleotide excision repair (NER) gene expression between monocytes and dendritic cells to support differences observed in functional repair assays.
    Status: Manuscript in preparation (contributing author).

  • Prognostic signature of DNA repair genes (TCGA)
    Builds a prognostic gene-expression signature from DNA repair genes using TCGA data; evaluates regularized survival models and overall survival.
    Status: Completed (course project).

  • Bayesian deconvolution of bulk RNA-seq with scRNA-seq references
    Applies a Bayesian deconvolution framework to infer cell-type composition and latent gene contributions in bulk RNA-seq using scRNA-seq references from human prefrontal cortex. Applied in a real-world AD dataset (e.g., GSE174367), adjusting for deconvoluted proportions improved AD risk prediction (AUC ↑ 0.641 → 0.705) and increased the APOE ε4 odds ratio by ~14.9%, suggesting composition confounding. Status: Completed (course project). Code accessible on GitHub repo.