Can the SNPbinner software be used for BC1 mapping populations?

[hurongimu]

Hi，
May I ask if the SNPbiner can be used for BC mapping populations?

cheers,
Rong

[Renqing1129]

Hi,
The same issue which i am trouble.

[itaygonda]

Hi. Sorry for my late response. We have never tried it for backcrosses. It works well for F2s. If you try to have input, please share.
Cheers
Itay

[hurongimu]

Hi,
I was very happy to receive your reply.Based on the analysis of the results generated by directly inputting the BC1 population data, I attempted the following approach, although I am not sure whether it is appropriate.
The BC1 population contains only h and b genotypes. I attempted to replace the h genotype with the a genotype, and in the resulting dataset, the a genotype intervals were merged with the intervals identified as h by SNPbinner and subsequently treated as h intervals.
The genetic map constructed using the bin markers generated from these data successfully reproduced one of our previously identified QTLs.

cheers
Rong

[itaygonda]

Thanks for sharing this info. I'm not sure it is appropriate as SNPbinner uses 2 different HHM, one for heterozygousema nd one for homozygote. Maybe @dauglyon will have a better insight, since he actually wrote SNPbinner code. Anyway, if that matches the QTLs you got with your previous work, maybe that's OK. The nice thing with QTL is that, theoretically, they should be unbiased analyses.
Best
Itay

[dauglyon]

I am not familiar with the b/h distribution in your data but.... this should be possible. It may require forking the code and manually changing the weights here (https://github.com/solgenomics/SNPbinner/blob/master/snpbinner/crosspoints.py#L40-L71) as h is intentionally underweighted. The code generally disfavors the h state so you may run into more wrinkles. Another option would be recoding your data as a/b to run SNPbinner, and depending on the distribution of these in the input, it may work.