Improve consensus partitioning via a hierarchical procedure

Consensus partitioning is an unsupervised method widely used in high throughput data analysis for revealing subgroups and assigns stability for the classification. However, standard consensus partitioning procedures are weak to identify large numbers of stable subgroups. There are two main issues. 1. Subgroups with small differences are difficult to separate if they are simultaneously detected with subgroups with large differences. And 2. stability of classification generally decreases as the number of subgroups increases. In this work, we proposed a new strategy to solve these two issues by applying consensus partitionings in a hierarchical procedure. We demonstrated hierarchical consensus partitioning can be efficient to reveal more subgroups. We also tested the performance of hierarchical consensus partitioning on revealing a great number of subgroups with a DNA methylation dataset. The hierarchical consensus partitioning is implemented in the R package cola with comprehensive functionality for analysis and visualizations. It can also automate the analysis only with a minimum of two lines of code, which generates a detailed HTML report containing the complete analysis.