correlation method of cluster annotation

18 views

Skip to first unread message

BigOmics Analytics Team

unread,

Aug 31, 2021, 5:04:04 AM8/31/21

to Omics Playground

I have a question regarding the correlation coefficient "R" mentioned in the cluster annotation under the section "clustering". I was wondering which correlation method was used to determine this coefficient?

[Thorben S., 19.08.2021]

BigOmics Analytics Team

unread,

Aug 31, 2021, 5:10:20 AM8/31/21

to Omics Playground

We use the Spearman's rank correlation for the calculation which is calculated as the Pearsons correlation on the ranks. For the genesets, we take the average relative log expression for the genes in the genesets as geneset expression value. Notice that p-value are not computed in this module. We could technically do that (computing p-value for correlation) but it would slow down the module considerably. For more traditional geneset enrichment and p-value, the user should use the geneset enrichment module under the "Enrichment" menu.