NYU Tandon Bioinformatics Symposium Showcases the Next Generation of Data-Driven Biomedicine

On April 25, the NYU Tandon Department of Biomedical Engineering convened a symposium that put a spotlight on an emerging truth in modern medicine: many of the most consequential discoveries now begin not at a lab bench or an operating table, but in the terabytes of sequencing data generated every day by hospitals and research centers around the world.

Organized by Mgavi Elombe Brathwaite, program manager, adjunct professor, and academic adviser for Tandon’s Bioinformatics Program, the event featured student researchers whose projects tackle diseases ranging from ovarian cancer to childhood esophageal disorders, all by asking, in one form or another, what the genome has been trying to tell us.

The breadth of the work on display underscored how rapidly bioinformatics has moved from a specialized subfield to a central research discipline. Cancer biology, inflammatory disease, orthopedic injury, autoimmunity, and chronic allergic inflammation were all represented. What unified the projects was a shared conviction that rigorous computational analysis of patient-derived molecular data can reveal patterns no single experiment could uncover.
 

A Personalized Approach to Crohn’s Disease

Destiny Collazo’s project, “Tailoring Crohn’s Disease Management through Molecular Fingerprinting,” addressed a question that every gastroenterologist eventually faces: why do two patients with the same diagnosis respond so differently to the same treatment? By profiling the molecular signatures unique to individual patients, Collazo’s work points toward a future in which Crohn’s management is guided not by trial and error but by a patient’s specific biological fingerprint—a shift that could meaningfully reduce the years some patients spend searching for a therapy that works.
 

Reading the Genome’s Hidden Passengers

Sara El Houzaly opened a window onto one of the genome’s least understood residents: LINE-1, a so-called “jumping gene” that can copy itself and reinsert elsewhere in the DNA. LINE-1 activity has long been suspected of contributing to the genomic chaos that defines high-grade serous ovarian cancer, but pinpointing exactly which copies are active (and where) has been nearly impossible with conventional sequencing methods. El Houzaly’s project uses long-read sequencing at single-cell resolution to identify specific, recurrent LINE-1 elements that appear across patients and across metastatic sites, and her findings suggest that a small subset of these elements may still be fully functional, potentially offering new targets for understanding how ovarian cancer evolves and spreads.
 

Small RNAs, Big Implications for Joint Injury

Jessica Flynn turned the spotlight to a different clinical problem: post-traumatic osteoarthritis, the form of joint degeneration that often follows ACL rupture and for which no disease-modifying therapy yet exists. Analyzing synovial fluid samples from patients with ACL injuries, Flynn identified dozens of microRNAs and PIWI-interacting RNAs—tiny regulatory molecules—whose levels shift meaningfully as the disease takes hold. These signatures could eventually serve as early diagnostic markers, or even as therapeutic targets to slow the cascade of damage before it becomes irreversible.

When Inflammation Leaves a Lasting Mark

Megha Lal’s contribution to the event involved research into eosinophilic esophagitis, a chronic allergic condition of the esophagus. Lal’s work focuses on IL-13, a cytokine central to type-2 inflammation, and asks whether its effects on the esophageal lining persist even after the inflammation subsides. Her findings suggest that IL-13 may leave a durable imprint on the epithelium—a kind of molecular memory encoded in chromatin—that helps explain why patients in apparent remission continue to experience barrier dysfunction. The implication is significant: effective long-term treatment may need to restore the epithelium itself, not just suppress the inflammatory signals that damaged it.

Envisioning Who Will Respond

Jehoiada Mebratu presented work aimed at one of the most frustrating problems in treating inflammatory bowel disease: a sizable fraction of patients simply do not respond to anti-TNF therapy, the current first-line biologic treatment, and others stop responding over time. Using stool-derived RNA sequencing data from a longitudinal cohort, Mebratu built a predictive model that distinguishes likely responders from non-responders at baseline, before treatment even begins. His analysis revealed that responders tend to show stronger epithelial repair signals, while non-responders are marked by amplified inflammatory signaling. If validated, the approach points toward a future in which a non-invasive test could spare patients from cycling through ineffective therapies.

Finding the Real Culprits in a Crowd of Suspects

Zoe Shtalryd took on a thorny statistical challenge at the heart of modern human genetics. When researchers identify genetic variants associated with disease risk, the true causal variant is often obscured by dozens of neighboring variants that travel together on the chromosome. Shtalryd is applying a fine-mapping approach across 28 immune cell types, using data from a large Japanese cohort that includes patients with ten categories of immune-mediated diseases. The goal is to move beyond broad associations and identify the specific variants that genuinely drive differences in gene expression—work that, despite its population-specific origin, is expected to have relevance far beyond the cohort studied.

A Burgeoning Program

Taken together, the presentations offered a compelling snapshot of NYU Tandon’s Bioinformatics Program and of the students Brathwaite and his colleagues are preparing for careers in academia, industry, and clinical research. Each project was grounded in a real clinical problem and pursued with the kind of methodological care that distinguishes meaningful bioinformatics from mere data-mining. Each also reflected something harder to quantify: the willingness to sit with complex, noisy, and often incomplete data long enough to find the signal inside it.

For the students who presented, the symposium was both a culmination and a beginning: a chance to share work that had occupied them for months or years, and a signal that their contributions are already being taken seriously by a community of mentors, peers, and faculty.

For those in attendance, it was a reminder that the future of medicine is being written, increasingly, in code.