Predicting Patient Survival from Longitudinal Gene Expression

Abstract

Characterizing dynamic gene expression pattern and predicting patient outcome is now significant and will be of more interest in the future with large scale clinical investigation of microarrays. However, there is currently no method that has been developed for prediction of patient outcome using longitudinal gene expression, where gene expression of patients is being monitored across time. Here, we propose a novel prediction approach for patient survival time that makes use of time course structure of gene expression. This method is applied to a burn study. The genes involved in the final predictors are enriched in the inflammatory response and immune system related pathways. Moreover, our method is consistently better than prediction methods using individual time point gene expression or simply pooling gene expression from each time point.

The role of longitudinal cohort studies in epigenetic epidemiology: challenges and opportunities

“…The ultimate goal of observational data generated in epidemiological investigations is to feed forward into clinical practice or public health. There is already evidence of translation of longitudinal biological data to clinical applications [ 83 ]. The incorporation of epigenetic biomarkers to enhance clinical tools for prediction and prognosis is beginning to emerge [ 5 ] (Table 2 ), and longitudinal cohorts will undoubtedly help in this domain.…”

Lagged principal trend analysis for longitudinal high‐dimensional data

“…Recently, dimension reduction and feature selection techniques were proposed and employed for longitudinal genomic data analysis. For instance, supervised learning methods were developed for time‐course gene expression data to predict patient outcomes (Zhang, Tibshirani, & Davis, ; Zhang & Ouyang, ) or classify patient subgroups (Zhang, Tibshirani, & Davis, ). Motivated by the problem of unsupervised learning of longitudinal genomic data, Zhang and Davis () proposed the principal trend analysis (PTA) method by integrating latent factor models for dimension reduction with spline‐based methods for temporal structure modelling (Wahba, ).…”

Joint Principal Trend Analysis for Longitudinal High-dimensional Data

“…Consequently, such characteristics raise challenges and opportunities on dimension reduction and feature selection of longitudinal genomic data. In supervised learning scenarios, time‐course structures of gene expression data have been used for predicting patient outcomes (Zhang et al, ; Zhang and Ouyang, ) and classifying patients (Zhang et al, ). In unsupervised learning scenarios, classical dimension‐reduction techniques, such as principal component analysis (PCA) and sparse PCA (d'Aspremont et al, ), have been widely used in static gene expression data analysis (Ringnér, ), combinational gene regulation modeling (Ouyang et al, ), and genome‐wide association studies (Aschard et al, ).…”