Methodological research
New measures and inference procedures for time-to-event data, including average hazard with survival weight, restricted mean survival time, and model-free group contrasts that go beyond the hazard ratio.
Dana-Farber Cancer Institute · Harvard Medical School
Statistical methods that make clinical evidence easier to interpret.
Uno1Lab develops and applies statistical methodology for biomedical research, with a particular focus on clinical cancer research — from new summary measures for survival data to tools for evaluating risk prediction models.
Treatment
Control
Schematic illustration, not study data.
The problem we work on
Beyond the hazard ratio
When two survival curves separate late — as they often do with immunotherapy — the treatment effect changes over time, and a single hazard ratio averages it away into a number that is hard to interpret clinically.
Our methods summarize the shaded difference directly. The average hazard with survival weight and the restricted mean survival time are model-free, stated in units clinicians and patients can reason about, and remain valid when proportional hazards does not hold.
What we do
Three strands of work
Methodological research, clinical collaboration, and open-source software — each one feeding the others.
Clinical collaboration
Collaborative investigations across oncology — study design, analysis, and interpretation — carried out with clinical investigators, geneticists, and data scientists at Dana-Farber and partner cancer centers.
Open-source software
Methods are released as documented packages — on CRAN for R and as installable commands for Stata — so that the procedures we publish can be used directly in real trials and observational studies.
About the lab
Led by Hajime Uno, PhD
Hajime Uno is Associate Professor of Medicine at Harvard Medical School and Dana-Farber Cancer Institute, working in the Department of Data Science and the Division of Population Sciences.
Our mission is to pioneer innovative statistical methods that enhance the understanding and interpretation of biomedical research results. Through rigorous methodological work and collaborative clinical investigation, we aim to advance evidence-based medicine and, ultimately, patient care and public health.