Related to pseudo test cross

22 views
Skip to first unread message

vska...@gmail.com

unread,
Jun 9, 2022, 9:24:03 AM6/9/22
to R/qtl discussion
Hello All,

We are utilising pseudo test cross strategy in plant population. I have extracted SNPs and checked for 1:1 segregation for (ABXAA) and 1:2:1 segregation for (ABXAB) loci. 
I have two queries:

1) I constructed a genetic map using loci with 1:1 segregation and using cross type "bc".
However, I don't see any correlation with physical positions. Has anyone also experienced similar thing. 

2) Is it possible to combine loci with 1:1 and 1:2:1 segregation and use for linkage map construction. I came across a paper wherein they mentioned that:

SNPtag data files were imported into Microsoft Excel and numerical values (0,1,2) were converted to A, H, B (representing AA, AB, BB). Segregation for each locus was obtained and tested for goodness-of-fit with predicted test-cross segregation (1:1) or F2-segregation (1:2:1) using chi-squared tests. Loci with significant chi-squared test statistics for goodness-of-fit for test-cross segregation (1:1) were combined with loci with significant test statistics for F2-segregation and subsequently combined into male and female data sets. These two data sets were imported as F2-formated files and separately analyzed in JoinMap 5.0 (VAN Ooijen, 2011).

Can we do that? If yes, is it possible to do it in Rqtl?

Thanks and regards

Sandip

genetic_physical_map_comparison copy.pdf

Karl Broman

unread,
Jun 10, 2022, 2:58:29 PM6/10/22
to R/qtl discussion

1) I constructed a genetic map using loci with 1:1 segregation and using cross type "bc".
However, I don't see any correlation with physical positions. Has anyone also experienced similar thing. 


If you have known physical positions, I wouldn't try to construct a genetic map from scratch, but would rather take the physical order and estimate the inter-marker distances. It's hard to tell without axis scales, but this could just be the limit of estimating marker order with the available data.
 
2) Is it possible to combine loci with 1:1 and 1:2:1 segregation and use for linkage map construction. I came across a paper wherein they mentioned that:

SNPtag data files were imported into Microsoft Excel and numerical values (0,1,2) were converted to A, H, B (representing AA, AB, BB). Segregation for each locus was obtained and tested for goodness-of-fit with predicted test-cross segregation (1:1) or F2-segregation (1:2:1) using chi-squared tests. Loci with significant chi-squared test statistics for goodness-of-fit for test-cross segregation (1:1) were combined with loci with significant test statistics for F2-segregation and subsequently combined into male and female data sets. These two data sets were imported as F2-formated files and separately analyzed in JoinMap 5.0 (VAN Ooijen, 2011).

Can we do that? If yes, is it possible to do it in Rqtl?

R/qtl doesn't have any particular functionality to aid with this.

karl
 
Reply all
Reply to author
Forward
0 new messages