Clusters plot
For the clusters plot, a hierarchical agglomerative cluster analysis is used to regroup PIP based on similarities. Briefly, in the initial stage, every PIP is placed into a singleton cluster. Then, the Pearson correlation coefficient is computed between every cluster and the two clusters with the highest correlation are merged into one. This step is repeated until the correlation between clusters is lower than a value set by the user (typically between 0.9 and 0.95). Singleton clusters are then removed. The clusters plot is a scatter plot representation of this hierarchical structure with, in x-axis the average centre of all the PIPs within the cluster and in y-axis the accurate mass of the most intense PIP. The size of the plot depends on the intensity of the most intense PIP. Clusters plot is obtained using
myPeakList.ClusterPlot(CC)
where CC is the cut correlation coefficient.
myPeakList.ClusterPlot(0.95)
Clusters plots provide a comprehensive representation of complex X-HRMS. As with the centergram representation, the data cursor tool can be used to attach to the graph key figure of merits and open a new window with information about the selected cluster. This is demonstrated below with the two clusters selected previously.
Grouping the PIPs by similarities not only allow the simplify the data representation but also to quickly recognise the molecular ions and its main isotopes drastically simplifying the identification of unknown species.
Up : Pure ion profile and peak list
Next :
Previous : Centergram representation
Related to: