Zsofia Sztupinszki

I am a physician-scientist with a background in oncology, cancer genomics, computational medicine, and artificial intelligence. I am currently a Research Fellow at Harvard Medical School and Boston Children's Hospital in the Computational Health Informatics Program, and an affiliate researcher at the Danish Cancer Institute.

My research focuses on using computational biology, machine learning, and AI to study cancer genomics and develop clinically relevant biomarkers. Over the years, I have worked on several areas of cancer research, including DNA repair deficiencies such as homologous recombination deficiency (HRD) and nucleotide excision repair (NER) defects, immuno-oncology, liquid biopsies, computational pathology, and the use of large language models in medicine and scientific discovery.

I developed scarHRD, an R package for estimating HRD from tumor sequencing data. It has been used and cited in more than 300 scientific publications. I also created TCGA Explorer, a no-code platform that allows researchers to analyse data across 30 TCGA cancer types without programming. My aim with these tools is to make cancer data analysis more accessible, reproducible, and useful for researchers and clinicians.

I have published 52 peer-reviewed papers, with more than 2,200 citations and an h-index of 25 on Google Scholar. I am also an inventor on patent WO2021195390A2.

I also write and make videos about integrating AI and large language models into real research workflows — from everyday LLM use and AI co-scientists to agentic coding — with a focus on doing so productively and responsibly. See Writing & Talks.