Header-only chromosome files when processing Geneseek

23 views
Skip to first unread message

Randall Marshall

unread,
Sep 25, 2020, 10:35:25 AM9/25/20
to R/qtl2 discussion
Hi Karl/all,

We're starting a new project looking for linkage between a fatal/nonfatal phenotype associated with an AJ mouse strain after a cross with NOD mice. We're eventually going to do Collaborative Cross mice, so we're using QTL2 to start to analyze GigaMUGA data.

We're using the Preparing DO mice tutorial to start. However when using the geneseek2qtl2.R script, it generates 21 chromosome files with headers only.

Any ideas?

Best and thanks!
Randall Marshall
Cincinnati Children's Hospital
Prows lab, Department of Human Genetics

finalreport header.jpg
genesqueakqtl2.txt
outputfilechr1.jpg

Karl Broman

unread,
Sep 25, 2020, 10:45:41 AM9/25/20
to R/qtl2 discussion
I think the header in the finalreport file is causing the problem. Try adding "skip=9" in the line
    g <- myfread(ifile)
to skip over that header. So it's like this:

g <- myfread(ifile, skip=9)
karl

--
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/cdf7be89-885d-4625-9285-89d4d7469ea4n%40googlegroups.com.

Randall Marshall

unread,
Oct 22, 2020, 4:11:38 PM10/22/20
to R/qtl2 discussion
I appreciate the help!! I added the skip=9
It goes through the finalreport.txt analysis (generating that box shape), and the issue (header only chromosome files) seems to stem from this error:

 -File: C:/Users/marzv4/Desktop/RQTL2/NODanalysis/finalreport.txt 
 -Reading data
|--------------------------------------------------|
|==================================================|
Error in `[.data.frame`(codes, , "marker") : undefined columns selected


Any additional thoughts or suggestions?
Best,
RAndall

Karl Broman

unread,
Oct 22, 2020, 4:46:20 PM10/22/20
to R/qtl2 discussion
I would step through the script line-by-line to isolate the line with the problem and then figure out the nature of the problem.

karl

Randall Marshall

unread,
Oct 22, 2020, 5:00:22 PM10/22/20
to R/qtl2 discussion
Good advice - you know what it was, it didn't like the header on the GM_allelecodes.csv file, either.
# read genotype codes
> codes <- myfread(codefile, skip=3)

fixed it..!

Off to discover new & exciting problems...

Best, RSM
Reply all
Reply to author
Forward
0 new messages