Immune responses span interconnected regulatory modalities within and across cells. Deciphering and targeting these multifactorial processes is still a major challenge, severely limiting our ability to harness the immune system, as well as other forms of multicellular regulation, for cancer treatment and prevention. To address this challenge, my newly established laboratory develops multidisciplinary engineering-based systems to study and manipulate multicellular circuits, aiming to (1) uncover molecular mechanisms governing immune recognition and function and (2) identify new immunomodulating interventions at an accelerated pace. To this end, we bring together latest advances in single-cell/spatial genomics, machine learning, and CRISPR-based genetic engineering to probe (multi)cellular circuits with greater resolution and scope via massively parallel experiments.

In this talk, I will describe our recent discoveries revealing new mechanisms controlling cellular and tissue immunogenicity, proposing epigenetic reprogramming and endogenous retroviral de-repression as a therapeutic modality to overcome immunotherapy resistance in cancer, and uncovering intercellular interactions controlling T cell recruitment and dysfunction. I will then describe our efforts in developing scalable systems to study and reprogram multicellular circuits, and how we have begun to utilize these systems to address fundamental questions in immunology and identify new immunomodulating and tissue-remodeling interventions.

In this talk, I will present an overview of our precision medicine program leveraging genomics and AI to deliver high-quality, outcomes-changing personalized medicine to cancer patients at Weill Cornell Medicine. I will describe our effort to bring whole-genome sequencing into the clinic, automate the interpretation of genomes and more broadly use whole-genome sequencing as a platform for AI-driven predictive medicine. I will also describe orthogonal techniques to obtain new insights into clinical samples, such as single-cell tissue imaging using imaging mass cytometry and describe the application of this technology coupled with analytics to turn millions of single cells into cancer subtypes that predict treatment response. In an effort to help augment the number of effective therapies targeting alterations we see in cancer patients, I will then discuss ongoing projects using AI to help develop novel therapies, focusing initially on predicting mechanisms of action of orphan molecules and positioning them in specific patient populations. I will finish by outlining some important future directions for AI in precision medicine.

Cellular plasticity, the ability to take on unexpected fates or identities, is a fundamental property of multicellular life that enables coordination during development and wound healing. Tumor cells expand this plasticity, which confers their capacity to adapt to new environments and survive therapy. My group uses single-cell genomics data from developmental systems and cancer models to understand the causes and consequences of cellular plasticity. I will describe approaches that we have developed for quantifying plasticity in the context of lineage transformation in prostate cancer and demonstrate how both oncogenic mutation and environmental insults can lead to dramatic increases in tumor cell plasticity. I will also highlight our work with models of pancreatic cancer initiation and progression, which suggests that epigenetic priming allows cancer cells to access diverse regenerative programs inappropriately, and renders them competent to participate in novel inductive signaling in the tumor microenvironment.

Prostate cancer is a common condition with limited treatment efficacy in castration-resistant metastatic disease, including with immune checkpoint inhibitors. I will discuss our efforts to apply single-cell RNA-sequencing to paired normal and cancer human prostate samples. We identified a previously unappreciated subset of tumour-enriched androgen receptor-negative luminal epithelial cells and a variety of innate and adaptive immune cells in normal prostate that were transcriptionally perturbed in prostate cancer. An exception was a unique, prostate-specific, zinc transporter-expressing macrophage population (MAC-MT), that contributed to tissue zinc accumulation in homeostasis, but showed enhanced inflammatory gene expression in tumours, including T cell-recruiting chemokines. Remarkably, enrichment of the MAC-MT signature in cancer biopsies was associated with improved disease-free survival, suggesting beneficial anti-tumour functions.