[R] how to quantify outliers on multi-dimensional scaling plot?

Ana Marija @okov|c@@n@m@r|j@ @end|ng |rom gm@||@com
Wed Mar 31 22:20:52 CEST 2021


I am in process of writing a grant where I am explaining my planned
methylation analysis using R software "minfi". In the text of the
grant I am mentioning looking for samples containing outliers in the
multi-dimensional scaling (MDS) plot
https://rdrr.io/bioc/minfi/man/mdsPlot.html . My question is: how to
more precisely define what is an outlier on a MDS plot? I know it is
not logFC like on volcano plot. Is there is a way to quantify distance
on MDA plot or it is just sufficient to say that samples that cluster
separately are outliers?

Here you can find a full workflow for minfi package including making MDS plots:


More information about the R-help mailing list