R/qtl generating an extensively longer linkage map comparing to Joinmap output

[moore...@gmail.com]

Hi Karl,

I am working on plant model species for a particular trait. Then, I constructed a linkage map from ~1000 SNP (selected 400K WGS SNPs). R/qtl is very efficient to deal with 400K SNPs for filtering poor SNP and clustering into linkage groups. 

However, it appears that the linkage map (from est.map()) tends to be extensively long (200 - 600 cM per linkage group). 

I decide to use Joinmap V4.1 to finalise the map with the same dataset. I took the same SNP to construct the map per linkage group. The map length is ~100 - 120 cM per linkage group. The difference between R/qtl and Joinmap linkage map is the order of loci.   

Does est.map() need a particular input parameter or 

Does est.map generate the map with a particular assumption? 

Best,

Jitpanu 

[Karl Broman]

est.map doesn’t reorder markers, but rather estimates inter-marker distances for a particular order. The function orderMarkers(j has a crude method for getting an initial order, But it needs to be followed by a lot of detailed manual work with ripple() and switch.order().

karl