DONDENA Seminar - Danilo Pellin

You may follow the seminar at the following link.

“Following cells after Gene Therapy: from genes to populations” 

SPEAKER: Danilo Pellin (Harvard University).

ABSTRACT:

Lentiviral gene therapy (LV-GT) enables durable correction of hematopoietic disorders by enabling the engraftment and long-term self-renewal of genetically modified hematopoietic stem cells. Ensuring efficacy and safety requires quantitative models that link genome-wide integration patterns to the long-term evolution of clonal composition. While integration site (IS) analysis provides rich longitudinal data, existing approaches typically focus on individual genes or dominant clones and do not capture system-level effects acting across the genome or over time. We introduce a two-component statistical framework for modeling IS data and clonal dynamics. First, we apply sparse fused lasso regression to genome-wide IS profiles to identify contiguous genomic regions exhibiting coherent enrichment or depletion, capturing distributed signatures of integration bias or selection without relying on predefined genomic features. Second, we model clonal composition using a Dirichlet-based formulation that treats clone abundances as a coupled multivariate system evolving over time. A simulation-based calibration strategy is used to assess uncertainty, characterize stochastic drift, and distinguish persistent selective effects from sampling variability. Together, these methods link spatial patterns of integration to long-term changes in clonal composition, enabling early detection of system-level perturbations and forecasting of clonal outcomes. The framework supports complex experimental designs across multiple cell types, time points, and therapeutic conditions and provides a principled foundation for comparative safety assessment in LV-GT and related gene and cell therapy applications.
 

BIO: 

I am an Assistant Professor at Harvard Medical School and a Principal Investigator in the Gene Therapy Program at Dana-Farber/Boston Children’s Cancer and Blood Disorders Center. My research group develops statistical and computational methods to study genome engineering tools and their outcomes, with a focus on clonal dynamics, promoter and protein engineering, and quantitative analysis of off-target editing. Our work aims to understand and predict how engineered genomic perturbations shape cellular systems over time, informing the design of safer and more effective gene and cell therapies.