IW-SRAHC: An Interactive Web-tool for Survival Risk Analysis based on Hierarchical Clustering

Abstract

In order to examine the effectiveness of selected features on the prognostic survival of patients, it is usually necessary to group the samples and compare differences between their's Kaplan-Meier curves. Traditional methods perform a hierarchical clustering analysis on expressions instead of using risk scores to classify samples into different risk groups, which may be effective only when there's only single-dimensional feature. Once multi-dimensional feature is provided, the clustering result is apt to lapse, for it is not themselves but their regression that is correlated to survival time. Besides, the dendrogram derived from the hierarchical clustering may lead to wrong classification. We propose an Interactive Web-tool for Survival Risk Analysis based on Hierarchical Clustering (IW-SRAHC), which contains a cohesive scheme including interactive sample division and automatic re-clustering using the hierarchical clustering results on risk scores. Kaplan-Meier survival curves are utilized as the measurement. IW-SRAHC is a highly automatic toolkit with no technical parameters required. It allows researchers to excavate more features consistent with survival time and associated with prognosis risks, and also provides a new way for researchers to validate their research findings. The software is freely available https://github.com/wylapp/iwsrahc.