Long and gappy map

[Siri B]

Dear Karl, 

We have made a genetic map for a F2 population of Arabidopsis lyrata (Austria x Norway) based on ddRAD data. The map remains very long after filtering for missing data, colocated markers, double crossovers, and markers with high segregation distortion, and there are big gaps in the middle of each linkage group. When comparing physical position with genetic distance, we get some very strange plots (see examples). Do you have the possibility to help us understand what is going on here?

I should add that Arabidopsis lyrata is outcrossing and initially we were planning to use a 4way cross type, but we did not have genetic data from the F1 individuals. We therefore decided to base our map on sites that were homozygous in the parents (approx 5000 markers). Perhaps this is making some strange patterns in our data? 

The plot of recombination fractions looks much better with markerlrt, but we were not able to use this info in ASMap (to which we import the data from rQTL). 

Thank you, 

Siri

[Karl Broman]

It's really hard to tell with these sorts of things. Do you see lots of tight double-crossovers (which could indicate a high rate of genotyping errors), or are there a few individual markers that are leading to the map expansion?

It looks like many chromosomes have a single very large interval. It could be that omitting markers where the parents aren't homozygous led you to throw out all of the markers in an interval. That wouldn't on its own result in map expansion, though the estimated lengths of these very long intervals could have very large standard errors.

karl

[Siri B]

Dear Karl, 

Thank you so much for your quick reply. After further analysis of our data, we suspect that some of the patterns we’re observing might be due to contamination or switched samples. Specifically, when looking at the genetic distance versus physical position plots (example given in the previous message), it seems as though two linkage groups may be anchored to the same chromosome (we used the anchor by chromosome option in ASMAP).

We sequenced F2s from three different crosses of the same species, and our PCA results indicate that some samples were either switched or mixed. However, even after cleaning up the data, the maps remain long and gappy, which is puzzling.

We’re at a bit of a crossroads with how to resolve this and would appreciate any insights or advice you may have. Interestingly, we also observed numerous tight double crossovers, but after discussions with the sequencing company, we’ve concluded that this likely isn't the primary issue. 

Best wishes,  

Siri