Our lab is currently working on both of these issues.
The target audience is plant biologists, but this is relevant for salmon as well. They review and discuss the many options available to discover and genotype SNPs in polyploids, including some of the tools we are using: Stacks, Freebayes, and SAMtools. They highlight the benefits of looking at the positive overlap between different approaches and end with some broad tool recommendations: SAMtools and Bowtie2, but also remind us that one size does not fit all.
In the Seeb lab we are lucky to now have multiple mapping families for most Pacific Salmon species and have started to construct consensus linkage maps that synthesize recombination events occurring in multiple parents (using MSTmap, MergeMap and LEPmap). While this is generally working well, none of these programs provide measures of uncertainly for the final marker ordering or spacing. This leaves our confidence in the final linkage map to be based on a mixture of experience and trial and error and very difficult to quantify.
The authors combine simulated data with known chromosome sequences to estimate the appropriate level of confidence for the ordering of marker triplets. They find the true marker spacing and family size affect the quality of the map, as expected. In particular, they find that markers within 2 cM are best estimated with family sizes >500. This is in general agreement with our results and informal understanding; on a large scale marker order is well established, but on the small scale, it is less so. Our lab’s consensus linkage maps often have many markers collapsed into a single position. Is this accurately reflecting our knowledge, or can we do better?