Error in .checkTypos(e, names_x) : Object 'gender' not found.

[kcardonal]

Hello!
I am trying to run the tool in my own dataset. When running the ASE with getGenicDP I tried to set the gender_file parameter following the documentation, so I created a text file with 2 columns the first contain the samples name as in the dt_anno file and the second column contains the corresponding gender for each sample (XX /XY) .However I got an error, besides when I check my dt_anno file (which is the output of AnnotateX) I find a gender column with all the samples labeled "female" this column was created authomatically, how can avoid/correct it? it is very important for the model to set the gender correct? or can I continue my analysis just ignoring the gender label?. How the model figure out cases in which for example you have sex chromosome aneuploidy, it still works for set XCI?

sorry for making so much questions in only one post...thank you for your help!

[SRenan]

Hi, Which version are you running? I removed all mentions of gender and replaced it with sex. Although I realize that I forgot to update the vignette. In addition, in more recent versions, I have changed getGenicDP to work without the sex column with the following warning:

"There are no 'sex' column in the dataset. If not all subjects are female, remove them or add a 'sex' column and run again."

So you can manually remove the males and run the command without specifying the sex column or the sex_file.
You can either get version 1.7.2 from this GitHub or get 1.7.0 from BioConductor devel..

The model does not detect or handle aneuploidies. While it is definitely important to the study of XCI, I made this to leverage the massive amount of RNA-Seq available for which most of the samples are XX.

I will be removing the last mentions of "gender" from the vignette. Thank you for reporting!

[kcardonal]

Hello! thank you for your quick response. I am using version 1.6 (from: https://bioconductor.org/packages/3.14/bioc/html/XCIR.html) I could try with any of the 1.7.X versions you suggest. However, i would like to ask you more questions before (sorry!). Did you mention that I can manually remove the males and run the command without specifying the sex column or the sex_file, also that you developed the tool to leverage on the hight amount of RNA-seq available from XX samples. So, that means I should not run the tool on male samples? In my case I am working with samples from healthy males XY and males with sex aneuploidies if I remove them I would have zero samples hahaha.
I had this doubt from the begging, because in your publication you clearly mention in the method description: " For a gene g in a female individual..." (so the model is based on the assumption of a female karyotype )but also you mention some comparative analysis at the gene expression level with male samples from Geudavis data set... So, you did detect the escaping genes in females with XCIR, and later, compare the gene expression of those genes between male and female samples, but the male samples where never processed in XCIR ... is this correct?

Thank you again, I am new in this field and your work has help me to understand and clarify some things in my mind!

[SRenan]

The tool infers XCI states in diploid samples. XCI just does not happen in males so you should indeed remove the males.

you did detect the escaping genes in females with XCIR, and later, compare the gene expression of those genes between male and female samples, but the male samples where never processed in XCIR ... is this correct?

That is correct. Males were used for DGEA as another way to asses XCI states to compare with XCIR results.