Factor analysis for survival time prediction with informative censoring and diverse covariates. (4th June 2019)

• Record Type: Journal Article
• Authors: McCurdy, Shannon; Molinaro, Annette; Pachter, Lior
• Is Part Of: Statistics in medicine. Volume 38:Number 20(2019)
• Page Start: 3719
• Page End: 3732
• Publication Date: 2019-06-04
• Subjects: diffuse lower‐grade glioma -- exponential proportional hazards -- factor analysis -- glioblastoma multiforme -- informative censoring -- integrative models -- latent variables -- lung adenocarcinoma -- lung squamous cell carcinoma; Medical statistics -- Periodicals; Statistique médicale -- Périodiques; Statistiques médicales -- Périodiques; 610.727
• DOI: 10.1002/sim.8151
• Languages: English
• ISSNs: 0277-6715

Abstract

Abstract : Fulfilling the promise of precision medicine requires accurately and precisely classifying disease states. For cancer, this includes prediction of survival time from a surfeit of covariates. Such data presents an opportunity for improved prediction, but also a challenge due to high dimensionality. Furthermore, disease populations can be heterogeneous. Integrative modeling is sensible, as the underlying hypothesis is that joint analysis of multiple covariates provides greater explanatory power than separate analyses. We propose an integrative latent variable model that combines factor analysis for various data types and an exponential proportional hazards (EPH) model for continuous survival time with informative censoring. The factor and EPH models are connected through low‐dimensional latent variables that can be interpreted and visualized to identify subpopulations. We use this model to predict survival time. We demonstrate this model's utility in simulation and on four Cancer Genome Atlas datasets: diffuse lower‐grade glioma, glioblastoma multiforme, lung adenocarcinoma, and lung squamous cell carcinoma. These datasets have small sample sizes, high‐dimensional diverse covariates, and high censorship rates. We compare the predictions from our model to three alternative models. Our model outperforms in simulation and is competitive on real datasets. Furthermore, the low‐dimensional visualization for diffuse lower‐grade glioma displays known subpopulations.