Endogenous retroelements (EREs) are viral-like sequences—molecular memoirs of ancient infections that now make up more than a quarter of the human genome. Our laboratory asks a simple question with big consequences: what do these sequences do to human cells today? We study how EREs, including human endogenous retroviruses (HERVs) and LINE transposons, shape gene regulation, cell identity, cell–cell communication, and responses to external and internal stimuli.

Most EREs are normally kept silent by epigenetic safeguards to protect genome integrity. Yet they are far from inert. Some have been repurposed by evolution as regulatory switches that fine-tune nearby genes; others can become disruptive when reactivated. Using a systems biology approach, we integrate single-cell multiomics, mathematical modeling, deep-learning frameworks, and functional genomics to map ERE activity across diverse human cell states from embryonic programs to immune and neural lineages to understand how ERE-driven regulation contributes to health and disease.

Our work is guided by two core questions:

1. How are EREs controlled, and when do they contribute positively to normal gene regulation, particularly during embryogenesis, hematopoiesis, and neurodevelopment?
2. Why do EREs “wake up” in stress and disease such as aging, hypoxia, viral infection, or chronic inflammation and how does this awakening reshape cellular physiology in conditions like neurodegeneration or autoimmunity?

Our research spans two complementary axes. First, we chart the genomic and epigenomic landscape of EREs during healthy development and homeostasis. At single-cell resolution, we build cell-type-specific barcodes of ERE–gene regulatory networks. We identify where EREs act in cis (as enhancers or promoters) and in trans (through regulatory RNAs and retroviral-derived products such as Gag, Env, or reverse transcriptase). We then use CRISPR-based perturbations to test causality, thus pinpointing which elements and regulators control specific cell states and functions.

Second, we investigate ERE activation in disease and stress contexts using primary human samples and relevant models. By combining spatial omics with deep learning, we trace aberrant ERE signatures in tissues and link them to immune dysregulation, tissue injury, and neuroinflammatory trajectories. This integrative strategy helps us connect ERE activation to specific cell types, niches, and disease mechanisms—moving from association to actionable hypotheses.

Ultimately, our goal is to illuminate a long-overlooked layer of genome regulation and translate it into insight: EREs can be both guardians and instigators in our DNA. By decoding when, where, and how they act, we aim to open new routes for biomarker discovery and targeted intervention in human disease.