GradientScanSurv—An exhaustive association test method for gene expression data with censored survival outcome

Abstract

Accurate assessment of the association between continuous variables such as gene expression and survival is a critical aspect of precision medicine. In this report, we provide a review of some of the available survival analysis and validation tools by referencing published studies that have utilized these tools. We have identified pitfalls associated with the assumptions inherent in those applications that have the potential to impact scientific research through their potential bias. In order to overcome these pitfalls, we have developed a novel method that enables the logrank test method to handle continuous variables that comprehensively evaluates survival association with derived aggregate statistics. This is accomplished by exhaustively considering all the cutpoints across the full expression gradient. Direct side-by-side comparisons, global ROC analysis, and evaluation of the ability to capture relevant biological themes based on current understanding of RAS biology all demonstrated that the new method shows better consistency between multiple datasets of the same disease, better reproducibility and robustness, and better detection power to uncover biological relevance within the selected datasets over the available survival analysis methods on univariate gene expression and penalized linear model-based methods.

ssGSEA score-based Ras dependency indexes derived from gene expression data reveal potential Ras addiction mechanisms with possible clinical implications

“…As expected, the ssGSEA scores derived from the Ras pathway genes showed a significant association with pancreas cancer patient survival outcomes in TCGA PAAD data ( Fig. 5), which is consistent with the initial observation that many individual Ras pathway genes have a significant association with patient survival outcome 18 . This shows that the aggregate scores derived from the ssGSEA method indeed may faithfully reflect the combined impact of individual genes if many of the genes already show significant association individually, www.nature.com/scientificreports www.nature.com/scientificreports/ indicating the power of the ssGSEA method and its potential broad application.…”

“…The significance of the overlaps between the gene lists and gene signatures was assessed using Fisher's exact test with customized R scripts. Survival analysis for association of ssGSEA scores of Ras-dependency related signatures or Ras pathway genes with cancer patients' survival outcome from TCGA was performed using an in-house analysis method described previously 18 . GeneGO database query and network analysis were performed using the GeneGO portal (https://portal.genego.com/cgi/data_manager.cgi#).…”

“…To explore the potential connection between the ssG-SEA scores derived from the Ras dependency signatures and cancer patient survival outcomes, the in-house survival analysis method GradientScanSurv 18 was applied to the TCGA data. Our method was demonstrated to have better performance than other methods for association assessment of continuous variables such as gene expression data with censored survival outcome 18 . The ssGSEA scores derived from the Ras dependency signatures are also continuous variables, which are suitable for such an association test with our GradientScanSurv method.…”

“…Additionally, using expression data to derive EMT signature-based ssGSEA scores for individual TCGA patient tumor samples, the same two pathways were identified, extending the applicability of RDIs to patient samples. Finally, the ssGSEA scores of RAS dependency-related signature and Ras pathway genes showed a significant association with cancer patients' survival outcome by our in-house survival analysis method 18 . Together, these observations indicate that computationally derived ssGSEA scores faithfully represent the levels of Ras dependency of cancer cell lines and cancer patient samples, directly providing insights for cancer research and potential clinical applications.…”

Gene Expression Patterns Distinguish Mortality Risk in Patients with Postsurgical Shock

“…In addition, survival analysis based on gene expression can also help to understand the outcome in terms of the underlying biology [24]. In this regard, the top four overexpressed genes identified in the current study were related to the immune system.…”