denovo map- scantwo

20 views
Skip to first unread message

nata...@gmail.com

unread,
Aug 11, 2021, 8:18:16 AMAug 11
to R/qtl discussion
Hi,
The scantwo() results of two LGs with interaction were as follows:
     Trait chr1 chr2 pos1f pos2f lod.full  pval  lod.fv1  pval  
1     Phe1  LG2 LG21    34    24 7.948200 0.246 6.592670 0.130 
2     Phe2  LG2 LG21    34    24 9.010371 0.262 7.528616 0.212 
  lod.int  pval pos1a pos2a  lod.add  pval   lod.av1  pval
6.1518923 0.088    30    40 1.796308 1.000 0.4407773 1.000
7.3046853 0.053    12    18 1.705686 1.000 0.2239308 1.000

Both LGs had no QTL evidence for any phenotype.
I wonder if this means that I should merge those LGs.
Should I?
Thanks a lot!

Karl Broman

unread,
Aug 11, 2021, 4:47:25 PMAug 11
to R/qtl discussion
Whether to merge linkage groups has to do with the associations among the genotypes at the markers, which is completely separate from what you learn from scantwo().
These scantwo results are saying that there look to be possible QTL on linkage groups 2 and 21 that show a potential epistatic interaction for these two phenotypes.

karl

nata...@gmail.com

unread,
Aug 12, 2021, 10:13:08 AMAug 12
to R/qtl discussion
thanks!
what is a good p. value for scantwo results?

ב-יום רביעי, 11 באוגוסט 2021 בשעה 23:47:25 UTC+3, ‪Karl Broman‬‏ כתב/ה:

nata...@gmail.com

unread,
Aug 12, 2021, 10:20:19 AMAug 12
to R/qtl discussion
Is there a connection between an interaction plot for a pair of loci to the two-qtl models (scantwo output)?
Both determine whether the pair is additive or epistatic.

ב-יום חמישי, 12 באוגוסט 2021 בשעה 17:13:08 UTC+3, nata...@gmail.com כתב/ה:

Karl Broman

unread,
Aug 12, 2021, 11:03:16 AMAug 12
to R/qtl discussion
See chapter 8 of the R/qtl book, https://rqtl.org/book

nata...@gmail.com

unread,
Aug 17, 2021, 5:27:25 AMAug 17
to R/qtl discussion
Ok thanks
What does the following scantwo output suggest (alpha = 0.5)? 

     Trait chr1 chr2 pos1f pos2f lod.full   pval  lod.fv1   pval   lod.int   pval
      phe3 LG19 LG20    34    32      8.8  0.051      6.1  0.281       3.9  0.995
 pos1a pos2a  lod.add   pval   lod.av1   pval
    34    32      4.9  0.338       2.1  0.700

Full, cond. interactive and additive models were significant.
I guess the additive is irrelevant as long as the cond. additive is not significant as well
thank you,
Nataly 

ב-יום חמישי, 12 באוגוסט 2021 בשעה 18:03:16 UTC+3, ‪Karl Broman‬‏ כתב/ה:

Karl Broman

unread,
Aug 18, 2021, 11:40:49 AMAug 18
to rqtl...@googlegroups.com
I would conclude that there’s insufficient evidence for a second QTL, as p=0.28 is bot small. Looking for small p-values for both full and fv1, or for both add and av1.

karl

On Aug 17, 2021, at 4:27 AM, nata...@gmail.com wrote:


--
You received this message because you are subscribed to the Google Groups "R/qtl discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rqtl-disc+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rqtl-disc/5eb5fa24-2e71-4fc5-9f83-a02f246774d3n%40googlegroups.com.

nata...@gmail.com

unread,
Aug 19, 2021, 5:35:31 AMAug 19
to R/qtl discussion
Thanks a lot!

ב-יום רביעי, 18 באוגוסט 2021 בשעה 18:40:49 UTC+3, ‪Karl Broman‬‏ כתב/ה:
Reply all
Reply to author
Forward
0 new messages