plotlodprofile without using refineqtl

[Jason Johns]

Hi,

We have decided to manually manipulate the location of a couple of QTL based on intuition. Essentially, we think there are 2 unlinked QTL (A & B) on the same chr. When I fit two QTL on that chr, QTL B (larger than A) gets split into 2 different QTL (B1 & B2) . We have biological reasons to believe that B1 & B2 are linked, and would only like to include A & B2 in our final model. However, when we run refineqtl, QTL A's position is refined to B1, effectively including B1 & B2 in the model. 

I'm trying to run plotLodProfile with the 'makeqtl' model I built with A & B2, but I cannot do so because I have to run refineqtl on them first, and refineqtl changes the model to B1 & B2.

Is there a workaround for this?

Thank you!

Jason

[Karl Broman]

I think you’ll have to just do this “by hand”. Loop over positions on the chromosome, call makeqtl to create a qtl object with varying positions for one QTL but the desired fixed position for the other, and then call fitqtl and extract the LOD score.

The function replaceqtl might also be helpful.

karl