qtl interval for multiple traits

[fishlike]

It's very convenient to do single-QTL scan and permutation for
multiple phenotypes all at once. I followed the example from the Rqtl
book to get the LOD score summary for every phenotype all together.
But how to get the QTL interval (lodint ()) for multiple traits?

Second question, what is the realistic limit of the number of traits
one can do all at once? I suppose this may depend on the computing
power of the computer. What kind of power and computing time is
required to do scanone and permutation for a couple of thousands
traits?

[Karl Broman]

> It's very convenient to do single-QTL scan and permutation for
> multiple phenotypes all at once. I followed the example from the Rqtl
> book to get the LOD score summary for every phenotype all together.
> But how to get the QTL interval (lodint ()) for multiple traits?

Try summary.scanone with format="tabByChr" or format="tabByCol"

> Second question, what is the realistic limit of the number of traits
> one can do all at once? I suppose this may depend on the computing
> power of the computer. What kind of power and computing time is
> required to do scanone and permutation for a couple of thousands
> traits?

The main problem one would run into would be memory. Running scanone with a few thousand traits should be no problem; the permutations would best be done on a cluster.

karl