I performed compareCluster() of clusterProfiler package with an identical dataset several times, and the results obtained were slightly different. I would like to know why this happens and how to avoid this. Thank you.