About

Elham Azizi joined Columbia University in 2020 as the Herbert and Florence Irving Assistant Professor of Cancer Data Research in the Irving Institute for Cancer Dynamics and an Assistant Professor of Biomedical Engineering. She is also affiliated with the Department of Computer Science, the Data Science Institute, and the Herbert Irving Comprehensive Cancer Center. Her multidisciplinary research utilizes novel machine learning techniques and single-cell genomic and imaging technologies to study the dynamics and circuitry of interacting cells in the tumor microenvironment. She holds a BSc in Electrical Engineering from Sharif University of Technology, an MSc in Electrical Engineering, and a PhD in Bioinformatics from Boston University. Her postdoctoral training was conducted in the Dana Pe'er Lab at Columbia University and Memorial Sloan Kettering Cancer Center.

Research topics

• Genetics
• Cancer research
• Biology
• Ecology
• Cell biology

Selected publications

• Immune Checkpoint Therapy Drives Maturation of a Cellular Neighborhood Nucleated by T Cell-APC Triads Enabling Spatially Compartmentalized Tumor Immunity

  bioRxiv (Cold Spring Harbor Laboratory) · 2026-04-12

  articleOpen access

  Spatially organized immune hubs of T cells and antigen-presenting cells (APCs) have been linked to immune checkpoint therapy (ICT) efficacy, yet the mechanisms underlying their function remain unclear. Using CODEX multiplex imaging, we longitudinally characterized the dynamic evolution of intratumoral cellular neighborhoods (CN) defined by triad interactions of CD4 and CD8 T cells with two distinct myeloid APC populations: cDC1s and IFN-gamma-activated macrophages. We termed this CN the immunity-promoting CN (IP-CN) and tracked its progressive development during tumor rejection induced by anti-CTLA-4/anti-PD-1 therapy. A coordinated IFN-gamma; and TNF-alpha signaling signature accompanied the IP-CN assembly. Over time, the IP-CN underwent functional maturation, forming specialized sub-neighborhoods that compartmentalized proliferating T cells at the tumor periphery versus cytotoxic T effector cells interacting with tumor cell targets. Our findings reveal a spatiotemporal mechanism by which the IP-CN sustains and amplifies cytotoxic T cell responses, demonstrating how T cell-APC neighborhoods orchestrate tumor immunity.

  PublisherOA PDFDOI

• Pain management after minimally invasive gastrectomy: do we need epidural analgesia?

  European Journal of Surgical Oncology · 2025-06-14 · 1 citations

  articleOpen access1st authorCorresponding

  INTRODUCTION: Epidural analgesia (EDA) has been the gold standard postoperative pain management in gastric cancer surgery for many years. As minimally invasive techniques have reduced surgical trauma, the necessity of EDA is being questioned. This study aimed to evaluate if minimally invasive gastrectomy without EDA was associated with shorter hospital stay and fewer complications compared to operations with EDA. METHODS: All patients undergoing minimally invasive gastrectomy for cancer at Karolinska University Hospital (Jan 2015-Aug 2023) were included and divided into groups operated with and without EDA. Data were collected prospectively in an institutional database and retrospectively from medical records. The primary outcome was the length of hospital stay. RESULTS: Among 211 patients, 43 (20.4 %) had EDA, and 168 (79.6 %) did not, with 12 (7 %) of the latter requiring EDA postoperatively. The median hospital stay was one day longer in group with EDA (median days 7 vs 6 p < 0.001), supported by multivariable adjusted analysis with a coefficient of -1.76 days (95 % CI: -9.01, -2.26 p = 0.003). Patients with EDA had lower pain scores (NRS 0-24h, 0 vs 2 p < 0.001) but required longer postoperative norepinephrine infusion (11.5h vs 0h p < 0.001) and one day longer stay in the high-dependency care unit (p < 0.001). Overall (55.8 % vs 39.9 %, p = 0.060) and severe surgical complications (20.9 % vs 9.5 %, p = 0.039) were more common in the EDA group. CONCLUSIONS: Minimally invasive gastrectomy without EDA can be associated with improved outcomes including shorter hospital stays, fewer postoperative complications, reduced postoperative norepinephrine use, and decreased high-dependency care duration.

  PublisherDOI

• AMICI: Attention Mechanism Interpretation of Cell-cell Interactions

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-09-24 · 1 citations

  preprintOpen accessSenior authorCorresponding

  Spatial transcriptomic data enable study of cell-cell communication, yet current analysis tools often fail to provide dynamic, interpretable estimates of interactions and their spatial range across tissue. We present AMICI, an interpretable attention framework that jointly estimates interaction length scales, adaptively resolves sender-receiver subpopulations, and links communication to downstream gene programs. AMICI recovers ground-truth interactions in semi-synthetic data, uncovers gene programs linked to cell communication in the mouse cortex, and reveals length-scale-dependent tumor-immune signaling that reinforces estrogen receptor (ER) programs in breast cancer.

  PublisherOA PDFDOI

• Towards Identifiability of Interventional Stochastic Differential Equations

  ArXiv.org · 2025-05-21

  preprintOpen access

  We study identifiability of stochastic differential equations (SDE) under multiple interventions. Our results give the first provable bounds for unique recovery of SDE parameters given samples from their stationary distributions. We give tight bounds on the number of necessary interventions for linear SDEs, and upper bounds for nonlinear SDEs in the small noise regime. We experimentally validate the recovery of true parameters in synthetic data, and motivated by our theoretical results, demonstrate the advantage of parameterizations with learnable activation functions in application to gene regulatory dynamics.

  PublisherOA PDFDOI

• Energy Guided Geometric Flow Matching

  ArXiv.org · 2025-09-25

  preprintOpen access

  A useful inductive bias for temporal data is that trajectories should stay close to the data manifold. Traditional flow matching relies on straight conditional paths, and flow matching methods which learn geodesics rely on RBF kernels or nearest neighbor graphs that suffer from the curse of dimensionality. We propose to use score matching and annealed energy distillation to learn a metric tensor that faithfully captures the underlying data geometry and informs more accurate flows. We demonstrate the efficacy of this strategy on synthetic manifolds with analytic geodesics, and interpolation of cell

  PublisherOA PDFDOI

• Interpretable Neural ODEs for Gene Regulatory Network Discovery under Perturbations

  arXiv (Cornell University) · 2025-01-05

  preprintOpen access

  Modern high-throughput biological datasets with thousands of perturbations provide the opportunity for large-scale discovery of causal graphs that represent the regulatory interactions between genes. Differentiable causal graphical models have been proposed to infer a gene regulatory network (GRN) from large scale interventional datasets, capturing the causal gene regulatory relationships from genetic perturbations. However, existing models are limited in their expressivity and scalability while failing to address the dynamic nature of biological processes such as cellular differentiation. We propose PerturbODE, a novel framework that incorporates biologically informative neural ordinary differential equations (neural ODEs) to model cell state trajectories under perturbations and derive the causal GRN from the neural ODE's parameters. We demonstrate PerturbODE's efficacy in trajectory prediction and GRN inference across simulated and real over-expression datasets.

  PublisherOA PDFDOI

• Author Correction: Mapping multimodal phenotypes to perturbations in cells and tissue with CRISPRmap

  Nature Biotechnology · 2025-03-05 · 1 citations

  erratumOpen access
  PublisherOA PDFDOI

• Domain-Invariant Feature Learning for Patient-Level Phenotype Prediction from Single-Cell Data

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-09-25

  preprintOpen accessSenior authorCorresponding

  Abstract Accurate prediction of patient-level disease status from single-cell RNA sequencing (scRNA-seq) data is critical to enabling precision diagnostics. However, study-specific artifacts induce spurious correlations that limit generalization and interpretability. We studied this problem in the context of Multiple Instance Learning (MIL), a framework where each patient is modeled as a set of single-cell profiles. To improve robustness to domain shifts, we propose an adversarial and metric-based approach that learns domain-invariant representations while preserving task-relevant biological variation. We benchmarked our method on a systemic lupus erythematosus (SLE) dataset with synthetically added spurious features and evaluated its performance on two real-world scRNA-seq atlases: a cross-tissue immune dataset and a COVID-19 severity atlas. Across all settings, we observed consistent improvements in out-of-domain accuracy and more biologically faithful model attributions. Our findings establish a new standard for robust, interpretable patient-level prediction under domain shifts using scRNA-seq.

  PublisherOA PDFDOI

• Squidiff: predicting cellular development and responses to perturbations using a diffusion model

  Nature Methods · 2025-11-03 · 16 citations

  articleOpen access
  PublisherOA PDFDOI

• Cellular states associated with metastatic organotropism and survival in patients with pancreatic ductal adenocarcinoma

  Nature Genetics · 2025-10-27 · 7 citations

  article
  PublisherDOI

Recent grants

• Integrative framework for identifying dysregulated mechanisms in the tumor-immune microenvironment

  NIH · $269k · 2018–2020

• Integrative framework for identifying dysregulated mechanisms in the tumor-immune microenvironment

  NIH · $652k · 2020–2023

Frequent coauthors

• Catherine J. Wu

  Harvard University

  51 shared

• Katie Maurer

  Dana-Farber Cancer Institute

  36 shared

• Dana Pe’er

  Memorial Sloan Kettering Cancer Center

  35 shared

• Robert J. Soiffer

  31 shared

• Jerome Ritz

  Dana-Farber Cancer Institute

  30 shared

• Pavan Bachireddy

  The University of Texas MD Anderson Cancer Center

  28 shared

• Cameron Y Park

  Columbia University Irving Medical Center

  24 shared

• Yinuo Jin

  Columbia University

  19 shared

Awards & honors

• NSF CAREER Award
• Tri-Institutional Breakout Prize for Junior Investigators
• NIH NCI Pathway to Independence Award
• American Cancer Society Postdoctoral Fellowship
• IBM Best Paper Award at the New England Statistics Symposium