> Say a gene has been localised to a particular chromosome by sib pair
> analysis. Without recombinants to define the region of "linkage", how can
> you determine which genetic markers flank the region of interest. Is it
> those that have positive p values, p values above a certain number, or
> something else?
It is the recombinants which define the linked region, but I suppose you
are saying you can't see them. You should perform a multipoint analysis
(e.g. with mapmaker-sibs) and obtain a likelihood statistic (MLS). The
likelihood statistic provides the measure of support for one place on
the map compared with another. You might have a region over which the
MLS is within 1 or 2 MLS units of the maximum and take this to be the
region most likely harbouring the locus. IMHO sib pair analyses
generally implicate quite wide regions.
Within a region, which markers have positive two-point p values depends
more on chance variation than real evidence for linkage.
