Survival Analysis on Rare Events Using Group-Regularized Multi-Response Cox Regression
Published in Bioinformatics, 2021
In this paper led by Ruilin Li, we describe a new method to fit a sparse Cox Model for multiple time-to-event phenotypes from a large-scale (> 1 million features) genetic dataset.
This is a product of collaboration with the Hastie and the Tibshirani labs and built on other techniques that we have developed in the Rivas lab.
- BASIL algorithm and
snpnet, J. Qian et al, 2020- This manuscript describes a method to obtain the exact solution for a large-scale penalized regression model (Lasso). We show several examples using quantitative traits and disease outcomes in UK Biobank.
- Single-response Cox model, Li et al, 2020
- This manuscript describes an extension of the BASIL algorithm to Cox model. We show examples using time-to-event phenotypes (onset of diseases) in UK Biobank.
- Sparse reduced rank regression, J. Qian et al, 2020
- This manuscript describes a method to fit large-scale penalized regression for multiple responses.
Reference: R. Li, Y. Tanigawa, J. M. Justesen, J. Taylor, T. Hastie, R. Tibshirani, M. A. Rivas, Survival Analysis on Rare Events Using Group-Regularized Multi-Response Cox Regression. Bioinformatics 37(23), 4437-4443 (2021). https://doi.org/10.1093/bioinformatics/btab095