Bayesian variable selection in a Cox proportional hazards model with the “Sum of Single Effects” prior

Yunqi Yang, Karl Tayeb, Peter Carbonetto, Xiaoyuan Zhong, Carole Ober, Matthew Stephens · 2025

Read Bayesian variable selection in a Cox proportional hazards model with the “Sum of Single Effects” prior by Yunqi Yang, Karl Tayeb, Peter Carbonetto, Xiaoyuan Zhong, Carole Ober, Matthew Stephens (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-06233.txt.

Description

Six statisticians extend a gene-hunting method to studies where the outcome is time until an event. The method, called SuSiE, picks out likely causal variants from dense genetic data by summing single effects, but it was built for simple outcomes. Fine-mapping is hard because nearby variants correlate almost perfectly and datasets hold hundreds of thousands of people with thousands of candidate variants.

The team adapts the prior to the Cox proportional hazards model, the tool for time-to-event data, and calls the result CoxPH-SuSiE. In simulations it beat Bayesian selectors, and on British biobank data it fine-mapped asthma risk. It named fourteen candidate variants in eight regions, six with strong evidence of being causal. Two are known to be harmful, and others sit in a stretch of DNA that regulates the GATA3 gene. The authors offer the method for survival studies.

Catalog record

CREATORYunqi Yang, Karl Tayeb, Peter Carbonetto, Xiaoyuan Zhong, Carole Ober, Matthew Stephens
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE74 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-06233

Browse the full library · About the Tower of Babel