Error in convert_sex...

15 views
Skip to first unread message

Jennifer Jacob

unread,
May 13, 2024, 12:15:20 PMMay 13
to R/qtl2 discussion
I am running a 'dof1' association and am getting the following error:

Error in convert_sex(control$sex, output$covar, control$sep, control$comment.char,  : 

  sex column "sex" not found in covar


Here is my control file script and the first few rows of the .csv file with the 'sex' column. 

All the mice are female. Is that a problem? If not, what am I missing? TIA!


Screenshot 2024-05-13 at 12.07.44 PM.png
Screenshot 2024-05-13 at 12.08.49 PM.png

Dan Gatti

unread,
May 13, 2024, 12:35:16 PMMay 13
to rqtl2...@googlegroups.com

Usually, this means that some file doesn’t have sample IDs aligned. I find that if my genotypes were ever a data.frame and if R changed the column names, that cryptic errors like this show up.

 

Dan

--
You received this message because you are subscribed to the Google Groups "R/qtl2 discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rqtl2-disc+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rqtl2-disc/5b0b9109-7ca7-4de7-afd7-776efc6f3648n%40googlegroups.com.

---

The information in this email, including attachments, may be confidential and is intended solely for the addressee(s). If you believe you received this email by mistake, please notify the sender by return email as soon as possible.

Karl Broman

unread,
May 13, 2024, 12:55:10 PMMay 13
to R/qtl2 discussion
I think include

covar_file=“dof1_covar.csv”,
crossinfo_covar=“ngen”

and delete the lines with crossinfo_file and crossinfo_codes

karl

Jennifer Jacob

unread,
May 13, 2024, 12:55:54 PMMay 13
to R/qtl2 discussion
Thank you Dan. I'll take a look at that.

Jennifer Jacob

unread,
May 13, 2024, 1:10:32 PMMay 13
to R/qtl2 discussion
Karl, I tried that and get this error instead; which was what led me to the code I posted previously.

Error in convert_cross_info(control$cross_info, output$covar, control$sep,  :
  1 missing values in cross_info (cross_info can't be missing).

Karl Broman

unread,
May 13, 2024, 1:11:31 PMMay 13
to R/qtl2 discussion
There is apparently a missing value somewhere in that ngen column. Find it and either fix it or remove that line.

karl

Jennifer Jacob

unread,
May 13, 2024, 1:16:15 PMMay 13
to R/qtl2 discussion
If I run the 'control_file' script as you indicated but designate the 'crosstype' as "do" rather than "dof1", it runs without errors.

On Monday, May 13, 2024 at 12:55:10 PM UTC-4 Karl Broman wrote:

Jennifer Jacob

unread,
May 13, 2024, 1:16:15 PMMay 13
to R/qtl2 discussion
Nevermind my last message, I'll remove the line.

Jennifer Jacob

unread,
May 13, 2024, 1:16:15 PMMay 13
to R/qtl2 discussion
I have an FVB in there, so that is where the issue probably lies. 
As a non-DO, what is the best way to indicate "ngen"?

Reply all
Reply to author
Forward
0 new messages