Hello!
I am currently performing for my first time a correlation analysis on my ChIP-seq dataset. I have used all my indexed .bam files as inputs to multiBamSummary
and used the .npz output as input to plotCorrelation
. I have generated a correlation heatmap trying first the Pearson and then the Spearman method.
The heatmap resulting from the Pearson method shows almost all extremely high scores (the vast majority in between 0.9 and 1), while the heatmap obtained from the Spearman method shows a broader range of scores (mostly ranging in between 0.6 and 1). I don't have a deeper knowledge of statistics and especially regarding correlation analysis. I would really appreciate some opinions about that and whether is preferred to use Person or Spearman method for correlation analysis of ChIP-seq datasets.
Thanks in advance! :)