Multi-Study Approaches: adventures from nutrition to genomics
Date: September 30th, 2022 @ 4pm-5pm
Location: Beaupre 105
Host: Gavino Puggione
Biostatistics and computational biology are increasingly facing the urgent challenge of efficiently dealing with a large amount of experimental data. High-throughput assays are transforming the study of biology, as they generate a rich, complex, and diverse collection of high-dimensional data sets. The increased availability of ensembles of studies on related clinical populations, technologies, and genomic features poses four categories of important multi-study statistical questions: 1) To what extent is biological signal reproducibly shared across different studies? 2) How can this global signal be extracted? 3) How can we detect and quantify local signals that may be masked by strong global signals? 4) How do these global and local signals manifest differently in different data types?
We will answer these four questions by introducing a novel class of factor analysis methodologies for the joint analysis of different studies. The goal is to separately identify and estimate 1) common factors reproduced across multiple studies, and 2) study-specific factors. We present different medical and biological applications, going from genomic to nutritional epidemiological data. In all the cases, we clarify the benefits of using our multi-study methods compared to the standard techniques.
Roberta De Vito is a statistician with a passion for developing statistical models and machine learning tools for cancer research and disorder risk, with a particular focus on nutritional epidemiology and genomics. Her main research interest is latent variable model, Bayesian nonparametric, variable selection via sparsity prior, machine learning and big data.
She completed her PhD in Statistical Science at the University of Padua, advised by Ruggiero Bellio and Giovanni Parmigiani at Harvard University, where she developed her thesis work.
Then, she was a postdoctoral research fellow at Princeton University in Barbara Engelhardt’s group, where she developed Bayesian and latent variable discrete models in high-dimensional biological data.
Currently, she is an Assistant Professor in the Department of Biostatistics and at the Data Science Initiative at Brown University.
She was recently awarded the Youden Award in Interlaboratory Testing for her “Multi-study Factor Model” paper.