Hello there, I was trying to label (Sex) the samples from ''GSE19784HOVON65'' data set using Y chromosome genes and I found a significant inconsistency between my labeling and the sample annotation that you provided. I mean, several female samples expressed Y chromosome genes.

Created by Ramyar Molania Ramyar
Hi Ram, Something does seem to be going in this dataset. I am looking into it more deeply and will get back to you. fyi
Thanks Ram, Can you give an of the patient ids. It will help as I explore this issue to make sure I am reproducing what you are seeing.
Dear Mike, If you just look at the expression of Y chromosome - genes such as: DDX3Y, EIF1AY and ZFY, you will see several female samples show vey high expression of those genes.
Dear Ram, Can you give an idea of the number of samples here. Generally I have found that cancer samples can be trickier in this regard compared to normal samples. I usually do a sex check using XIST for example, but because cancer cells can have some stem like qualities XIST can actually get shut off. I have even seen XIST on in a male sample that had some chr X duplication (but that is rare). Additionally, there are homologs between some chr x genes and chr y genes that can yield some level of expression. Do you see many chr y genes on in these samples? and do these samples have XIST off or on? Although XIST is a long none coding RNA, the hgu 133 plus array does have probes for it. I have listed them below probe_id symbol 214218_s_at XIST 221728_x_at XIST 224588_at XIST 224589_at XIST 224590_at XIST 227671_at XIST 235446_at XIST 243712_at XIST If these samples have both exit off *and* multiple chr y genes expressed please let me know. Here is a [paper](https://www.ncbi.nlm.nih.gov/pubmed/20556862) that may be of interest.

questions about 'GSE19784HOVON65' data set page is loading…